JPH07202906A - Network connector and information communication system - Google Patents

Abstract

PURPOSE:To connect high-speed networks by providing a connection 1 setting/ managing means for managing the set ATM connection and a protocol converting means for converting a protocol. CONSTITUTION:Two of different LAN 111 and 112 are connected by using the network connector (IWU), IWU 221 and 222 are respectively provided at the LAN 111 and 112, an ATM protocol is adopted between the IWU 221 and the IWU 222 as an original protocol between the IWU, the protocol conversion between the LAN 111 and the ATM protocol is performed at the IWU 221, and the protocol conversion between the LAN 112 and the ATM protocol is performed at the IWU 222. On the other hand, inside the IWU 221, the connection managing means exists for managing and controlling the connection while making the connection between the IWU 221 and 222 correspondent to the destination information of information transmitted from a terminal inside the LAN 111.

Description

Detailed Description of the Invention

[0001]

FIELD OF THE INVENTION This invention relates to network systems.

[0002]

[Prior art]

<1> As a network connection device in a communication network system such as a conventional LAN, a bridge, a router, a gateway, etc. are provided. The bridge is a network connection device having only a so-called address filter function, such as terminating and processing up to layer 2 of the communication protocol between networks. The router has a slightly higher function than the bridge, and terminates and processes up to layer 3 of the communication protocol. The gateway is capable of terminating up to layer 7 of the communication protocol and completely converting the protocol for network connection.

However, such a conventional network connection device can only provide one-to-one protocol conversion by software processing. Therefore, when connecting from one LAN to multiple LANs,
It is necessary to separately prepare a network connection device for performing protocol conversion to each of them, which increases the amount of required hardware and makes it difficult to take a flexible network configuration. It was

On the other hand, in recent years, a network connection device such as a multi-protocol router capable of supporting a plurality of protocols has been proposed or put into practical use, and a plurality of networks are connected by one network connection device. .

However, similarly to the above-mentioned network connection device such as a gateway, software processing is performed to perform the protocol processing of the input network up to the upper layer and convert it to the protocol of the corresponding network. After that, network connection was performed by a method of performing protocol conversion to a predetermined lower layer.

Such protocol conversion by software processing is not only easy to implement, but also easy to implement in a high-performance network connection device such as a multi-protocol router, so that the current network connection device can be implemented. It is a method that is often used in protocol conversion in.

However, there has been a problem that protocol conversion by software processing takes much longer than usual. In particular, in a multi-protocol router that processes multiple protocols at the same time, a considerable amount of software processing is required, so when network connection is performed between networks performing high-speed data communication, The decrease in throughput due to such software processing has been a serious problem.

In particular, a connection-type high-speed network such as the ATM system and a LAN currently in operation
A network connection device for connecting to a datagram communication network such as Ethernet or FDDI may need a function of simultaneously performing conversion processing of a plurality of protocols at high speed and a function of performing network identification at high speed. It is expected that it is difficult to perform such high-speed processing in the current network-connected device using software processing.

<2> On the other hand, in the corporate communication network (LAN), in recent years, high speed, multimedia, large scale,
The need for high reliability is increasing. In particular, there is a strong need for speeding up "to realize higher speed information transfer", and research and development of an ultra-high speed LAN having a transmission speed of several hundred Mbps and several Gbps class has been actively conducted.

Regarding the explanation of LAN technology in general, the prior art is described in the literature such as "Information Network Handbook" (edited by the Institute of Electronics, Information and Communication Engineers), but the LAN system form is roughly divided. In the computer field, shared-me, in which multiple terminals are connected by a shared transmission line (bus, ring)
Switch-based star LAN with high-speed interface by combining dia system and communication field flow
And, it is classified into.

FIG. 39 shows a typical shared-med
FIG. 1 is a diagram showing a configuration of an ia LAN system, which includes a plurality of terminals 2002 and 2004 and shared transmission paths 2001 and 2003 connecting these terminals, and each terminal 20
02,2004 is a specific protocol such as CS
According to the access control method such as the MA / CD method and the token method, access competition is avoided for the transmission paths 2001 and 2003, and information is transmitted and received between the terminals.

The shared-media type LAN is a form in which control is distributed and routes are centralized. Since a plurality of terminals share a transmission line (bus, ring), the number of terminals increases as the number of terminals increases. However, there is a problem that access competition of the shared transmission line increases, and the shared transmission line becomes a throughput bottleneck. Especially when the system becomes large and the throughput of the whole network becomes large,
The ed-media type LAN requires an ultra-high-speed transmission line corresponding to the throughput of the entire network. Therefore, in order to increase the speed of the shared-media LAN, it is necessary to solve the problem that the shared transmission line (connection link) becomes a bottleneck.

FIG. 40 shows a typical star LAN.
It is a figure which shows the structure of a system, The some terminal 2006
And a switch 2005 accommodating a plurality of terminals 2006, each terminal 2006 sends a packet to the switch 2005 by using an individually given transmission path 2007, and the switch 2005 sends the packet from each terminal 2006. The switch 2 receives the packet and transfers the information to a desired output port based on the destination information of the packet or the routing information created based on the destination information of the packet.
When the transfer destination terminal receives the packet transmitted from 005, information is transmitted and received between the terminals.

The star system LAN is a form in which control is centralized and routes are distributed, and each terminal is individually provided with a transmission path, and there is no access competition for the transmission path of each terminal. Since the throughput of the transmission line does not decrease as the number of terminals increases, the network configuration is suitable for speeding up.

However, also in the star system LAN, when the system is scaled up and a plurality of nodes 2008 to 2011 are connected to construct an entire system as shown in FIG. 41, a connection transmission line (transmission between nodes) in the system is constructed. Roads 2012-2015 share information transfer from a plurality of terminals 2016, so that as the number of terminals increases,
The information transfer capability of the inter-node transmission line becomes a bottleneck, which causes a problem that the throughput is not improved. Therefore, even when speeding up the star LAN, sh
It is necessary to solve the problem that the information transfer capability of the inter-node connection link becomes a bottleneck, which is the same problem as in the ared-media type LAN.

Conventionally, in order to solve the problem that the information transfer capability of the inter-node connection link becomes a bottleneck, (1) a method of improving the information transfer capability itself of the inter-node connection link (2) inter-node A method is used to effectively utilize the information transfer capacity of the connection link.

In order to improve the information transfer capability itself of the inter-node connection link, a method of increasing the number of connection links as shown in FIG. 42 or increasing the speed of one connection link as shown in FIG. 43 ( The method of changing the link speed between nodes to static (semi-fixed) is adopted.

Further, in order to effectively utilize the information transfer capability of the inter-node connection link, as shown in FIG. 44, a method of dynamically changing the inter-node connection link speed as needed (variable) Link speed method) and Fig. 45
As shown in, a method of using multilink transfer, which is a transfer method using multiple paths at the data link level, has been adopted.

As described above, various studies have heretofore been made to improve the link speed between nodes and to effectively utilize the information transfer capability of the link between nodes in the information communication system, but at the data link level. There are many studies, and in a system configured by connecting multiple (three or more) packet processing nodes due to large scale, etc., there is limited study of "what is the physical network topology in which bottlenecks are unlikely to occur?" Only the network configuration has been studied (the network configuration that can effectively utilize the in-system information transfer capability at the network level has not been studied much).

Here, the data link level means information transfer between adjacent packet processing nodes (information transfer in which information is transferred between the same buffer queues when transferring packet information, and a plurality of transfers are performed). Allow to transfer using the route).

The network level is End-
It means the information transfer of the packet processing node between End-End when a plurality of packet processing nodes exist between the packet processing nodes between Ends (when transferring the packet information, through different relay queues). It means that information transfer will take place).

However, as the scale of the network becomes larger, the information communication system has a configuration in which nodes (2020 to 2024) are complicatedly connected as shown in FIG. There is. In such a multipath environment, when information is transferred, the total remaining bandwidth of the inter-node connection links is larger than the required bandwidth of the user ATM terminal, and the remaining bandwidth of any one connection link is the user ATM. It may be smaller than the required bandwidth of the terminal, and in this case, although the information transfer capacity of the inter-node connection link of the system is not insufficient,
A problem arises in that the bandwidth required by the user ATM terminal cannot be satisfied and allocated.

For example, as shown in FIG. 47, I / 2
The shortest route 203 for the user ATM terminals D and E of [Mbps] toward the user ATM terminals C and A, respectively.
With 2,2031 connections set, I
ATM terminal F for user of [Mbps] is A for user
When trying to establish a connection to TM terminal B,
In both cases of the transfer paths 2031 to 2034 and 2032 to 2033, only the maximum I / 2 [Mbps] transfer capacity can be assigned, and the connection link becomes a bottleneck. Actually, I / 2 [Mbps] still remains. Despite the remaining bandwidth, it cannot be allocated.

In order to solve this problem, allocation of inter-node connection resources is not performed for each physical link, but bandwidth allocation over a plurality of physical links is allowed, and information is transmitted through a plurality of physical links. It can be solved by making a transfer.

That is, the transfer paths 2031 to 2034,
Both 2032-2033 are used to calculate the total I [Mb
This is a method of transferring information in the band [ps]. However, even with this method, as shown in FIG.
31-2034 and transfer paths 2032-2033 are queued in different nodes 2041 and 2043, respectively, and then transferred to B. Therefore, in addition to transfer path transfer delay,
The difference in the queuing delay within the node also poses a problem that the order of information transfer is reversed.

Conventionally, the problem of reversal of information transfer order in such multi-path transfer is that the nodes receiving information from a plurality of routes rearrange the information in order, that is, the information transferred earlier is multi-path transferred. A method is used in which the transmission sequence numbers are waited for, and when the transmission sequence numbers are completed, information is transferred to the next node.

However, in the case of such a method, depending on the internal state of the network (the number of nodes passing through, the queuing state, etc.), it is better to perform multipath transfer when high-speed transfer is possible and to perform single path transfer. There is a problem that high-speed transfer is not possible and the information transfer capacity of the network in the multipath environment cannot be effectively utilized no matter which transfer method is used. This means that even if the processing capacity of system components (nodes, links) is sufficiently high, the processing capacity of the nodes and the transmission capacity of the links may not be effectively utilized in the entire system in which multiple nodes are connected. It means that there is.

The problems of the prior art described above can be summarized as follows: (a) In an information communication system configured by connecting a plurality of packet processing nodes, the inter-node information transfer capability of the packet processing nodes is max (node (Remaining bandwidth of inter-node connection link) <band required for inter-node information transfer <Σ remaining bandwidth of inter-node connection link, the single-path transfer is sufficient for the inter-node information transfer capability of the packet processing node. Nevertheless, the throughput will decrease.

(B) In order to solve the above problem a, when a method is adopted in which bandwidth allocation over a plurality of physical links is allowed and information transfer is also allowed through a plurality of physical links, Inversion of the information transfer order occurs.

(C) Depending on the internal state of the network (the number of nodes passing through, the queuing state, etc.), the case where multipath transfer is advantageous for high-speed transfer and the case where single path transfer is advantageous for high-speed transfer will change.

[0031]

[Problems to be Solved by the Invention]

<1> One LAN in the conventional network connection device
When connecting to multiple LANs, it is necessary to individually prepare a network connection device for performing protocol conversion to each LAN, which increases the amount of required hardware and provides a flexible network configuration. There was a problem that it became difficult.

Further, in the conventional network connection device using the protocol conversion by software processing so as to be compatible with a plurality of protocols, there is a problem that the processing takes a very long time. Especially, in the case of processing multiple protocols at the same time, a considerable amount of software processing is required, so throughput is reduced when network connection is made between networks performing high-speed data communication. Was a big problem.

The present invention has been made in view of the above circumstances, and makes it possible to realize high-speed inter-network connection for a plurality of connected networks, and to provide a configuration independent of the configuration and characteristics of the connected networks. Can connect to multiple types of networks with the same network connection device, and even when connecting multiple types of networks, it is possible to connect to a network that is unaware of the type of network being connected. It is an object of the present invention to provide a network connection device that can be easily designed and can significantly reduce the types of protocol conversion functions required.

Further, the present invention guarantees the same communication service quality as that provided in the ATM network even in the communication path between the network connection devices.
By providing, it aims to provide a communication service between networks that is most suitable for the resource status in the communication path between network connection devices.

<2> On the other hand, in both the shared-media LAN and the star LAN, not only the speedup of the system components (nodes and links) but also the information transfer capacity of the inter-node connection link is bottlenecked. The problem that becomes the bottleneck is solved, and the system components (node,
In order to improve the throughput of the entire system configured by connecting (links), it is an important technical issue to effectively use the processing and transmission capabilities of system components (nodes, links).

The present invention has been made in view of the above problems, and system components (nodes, links) are provided in order to improve the throughput of the entire system configured by connecting system components (nodes, links). It is an object of the present invention to provide an information communication system capable of effectively utilizing the information transfer ability of the above.

[0037]

[Means for Solving the Problems]

<1> (Invention 1) In the present invention (Claim 1), in each of the network connection devices provided in each of a plurality of networks for transmitting communication information via an inter-network communication path, each of the network connection devices is Connection setting / management means for setting an ATM connection to the inter-network communication path between the network connection devices and managing the set ATM connection, and communication between each network and the inter-network communication path And a protocol conversion means for performing protocol conversion between the predetermined protocol used in the network and the ATM protocol used in the inter-network communication path for the communication information to be transmitted. And

(Invention 2) In the present invention (claim 2), in each of the network connection devices provided in each of a plurality of networks for transmitting communication information via a plurality of inter-network communication paths, each of the network connection devices is , Connection setting / management means for setting an ATM connection to at least one of the inter-network communication paths between the network connecting devices and managing the set ATM connection, and between each network and the network. Protocol conversion means for performing protocol conversion between the predetermined information used in the network and the ATM protocol used in the inter-network communication path for the communication information communicated between the communication paths; This protocol conversion means and the plurality of networks A switch unit connected to the inter-workcommunication path, and based on destination information added to the communication information transmitted from each of the networks to the inter-network communication path, from among the plurality of inter-network communication paths, And a communication path selecting means for selecting an inter-network communication path to be connected to the protocol converting means.

(Invention 3) Further, in a network connection device provided corresponding to a network group composed of a plurality of networks and connecting three or more corresponding network groups by mutually connecting using a plurality of communication paths. A protocol used in the network and an ATM protocol used in the communication path, which is provided between the network group and the communication path, and which corresponds to information input from any one of the networks in the network group and the communication path. Protocol conversion means for performing protocol conversion between the communication paths and the network, and a plurality of networks forming the network group and ATM switches connected to the communication paths, the network group Included in the information sent from one of the Communication path selecting means for selecting the communication path to which the information is to be transmitted from the destination information to be transmitted, and selecting a network to which the information is to be transmitted from the destination information included in the information transmitted from any one of the communication paths. And a connection setting / management unit for setting an ATM connection with another network connection device using the communication path and managing the set ATM connection.

(Invention 4) Further, the above invention 1 (claim 1).
Alternatively, the invention 2 (claim 2) may further include a failure detection / avoidance means for detecting and avoiding a failure in the communication path between the network connection devices and the network connection device itself. Preferably, the failure detecting / avoiding means is a means for detecting a failure in the communication path between the network connecting devices, and a failure status in the communication path between the network connecting devices is stored in the network connecting device. Failure information notifying means for notifying the connection setting means, failure avoiding means for resetting the connection between network connecting devices according to the sent failure information, and connection between network connecting devices even if failure avoidance is performed A connection management means for managing whether or not there is an influence, and OAM
It may be configured to have a resource management unit that rewrites the resource information in the communication path between the network connection devices according to the OAM information from the function providing unit.

(Invention 5) In the present invention (Claim 3), the ATM connection set in the inter-network communication path in the above Invention 1 (Claim 1) or Invention 2 (Claim 2) is It is characterized by further comprising a communication service quality providing means for providing a predetermined communication service quality.

Further, the communication service quality providing means is
When a connection is set up between network connection devices, a communication service quality information collecting means for collecting information about the communication quality of service requested by the connection, and a connection set between the network connection devices are requested. The communication service quality storage means for storing the communication service quality, the communication service quality monitoring means for monitoring the communication service quality of the connection set in the network connection device, and the connection between the network connection devices are set. It may be configured to have a communication service quality guaranteeing means for guaranteeing the communication service quality, and a communication service quality managing means for managing the processing relating to the communication service quality provided between the network connection devices.

(Invention 6) In the above Invention 5, the communication service quality providing means may be a connection type connectionless connection type and a connectionless type connection type among the ATM connections set between network connection devices. It may be characterized in that different communication service qualities are provided for the connections performing the communication.

Preferably, the communication service quality providing means sends information to the network connection device to the network connection device and information identification means for identifying whether the information sent to the network connection device is information requesting the communication service quality or not. A combination of the communication service quality determining means for determining the communication service quality to be set in the connection between the network connecting devices according to the type of information coming in, and the communication service quality set by the communication service quality determining means and the connection for setting the communication service. It may be configured to include a communication service quality storage unit that stores the information, and a unit that provides the communication service quality set for the connection between the network connection devices.

(Invention 7) Further, the above Invention 5 or Invention 6
In the communication service quality providing means, the communication service quality providing means provides A between the network connection devices performing connection-type communication.
It may be characterized in that the TM connection is provided with a communication service quality higher than the communication service quality that is always required.

Preferably, the communication service quality providing means stores the required communication service quality information when the connection setting request information is sent to the network connection device, and the required service. A communication service quality determining means for determining the communication service quality to be set in the connection between the network connection devices corresponding to the setting request connection from the quality,
Communication service quality storage means between network connection devices for storing correspondence between the determined communication service quality and the connection for which the communication service quality is set, and means for providing the determined communication service quality to the connection between the network connection devices It may be configured to include and.

(Invention 8) Further, Invention 5 or Invention 6 described above.
In the communication service quality providing means, the communication service quality providing means provides A between the network connection devices performing connection-type communication.
It is characterized in that a communication service quality lower than the required communication service quality is provided to the TM connection according to the resource condition of the communication path between the network connection devices in which the ATM connection is set. Is also good.

Preferably, the communication service quality providing means stores the required communication service quality information when the connection setting request information is sent to the network connection device, and the required service. Communication service quality determining means for determining the communication service quality to be set in the connection between the network connecting devices corresponding to the setting request connection from the quality and resource information between the network connecting devices, and the determined communication service quality and its communication service quality May be configured to have a communication service quality storage means between network connection devices for storing the correspondence with the set connection, and a means for providing the determined communication service quality to the connection between the network connection devices.

(Invention 9) Further, Invention 5 or Invention 6 described above.
In the above, the communication service quality providing means provides different communication service qualities depending on the characteristics of the connectionless type information sent to the ATM connection between the network connection devices which are performing the connectionless type communication. May be featured.

The communication service quality providing means, when the datagram communication information is sent to the network connection device, the information identifying means for identifying that the information is the datagram communication information, and the sent datagram. Datagram communication information characteristic reading means for reading characteristics such as packet length and burst length of communication information, and communication service quality storage means for storing communication service quality of a connection between network connecting devices corresponding to the read characteristics, And a communication service quality determining means for determining a communication quality of service to be set in the connection between the network connecting devices which carries the information based on the characteristics of the datagram communication information, and to provide the set communication quality of service to the connection between the network connecting devices. It may be configured to have means.

<2> (Invention 10) On the other hand, in the information communication system of the present invention, the information transfer capability can be reassigned to a plurality of physical links at the data link level, and mixed accommodation of multiple speeds is possible. The packet processing node described above and transmission means for interconnecting the packet processing nodes are provided, and a single packet processing node is provided for all the other packet processing nodes at the network level of the physical link. It is characterized by connecting with a path.

That is, in an information communication system having three or more packet processing nodes connected, when the packet processing nodes are connected, the existence of a plurality of physical link connections at the data link level is allowed, but the data link level is allowed. The packet processing node is connected so that there is no multipath at the network level when the system configuration is replaced by replacing a plurality of physical links in 1. Of the information processing capability of the packet processing node that can be re-assigned and can accommodate mixed speeds of multiple speeds, and the All information transfer processing capabilities other than the information transfer capability of the processing node are Prepared (reserved) as the information transfer capacity for inter-connection links, and allow the information transfer capacity for inter-node connection links to mix multiple speeds in one (or more) physical links connected between packet processing nodes. However, it is possible to allocate as much as possible and perform bandwidth allocation when transferring information between the packet processing nodes based on the total remaining bandwidth of one or more physical links existing between the packet processing nodes. Therefore, it is possible to allow information transfer having a transfer band larger than the maximum remaining band of the physical link existing between the packet processing nodes, or to increase the information transfer capacity assigned to a certain physical link. It is characterized by

Further, in the above-mentioned configuration, there is provided a means for rearranging the information received from a plurality of routes in order so that the information can be transferred to the next node when the transmission order numbers of the multilink are aligned. good.

(Invention 11) In the information communication system of the present invention (Claim 4), a plurality of packet processing nodes for exchanging packets with at least one terminal accommodated therein and for exchanging the packets, and the packet. Transmission means for interconnecting processing nodes, wherein each of the packet processing nodes has a combination of the packet processing nodes that relays the packet for each of the packet processing nodes that are the transfer destinations of the packets, and The feature is that the order of relaying packets is uniquely determined.

That is, in the information communication system constituted by connecting three or more packet processing nodes, the existence of a plurality of logical link connections at the data link level is allowed when logically connecting the packet processing nodes. But
When multiple logical links at the data link level are replaced with one logical link and the system configuration is examined, a logical network is constructed, operated and managed so that there are no multipaths at the network level, and packet processing is performed. Of the information transfer capacities of the nodes, all the information transfer processing capacities other than the information transfer capacities of the packet processing nodes allocated to accommodate terminals are prepared (reserved) as the information transfer capacities for connection links between packet processing nodes, and the node The information transfer capability for the inter-connection links is allocated to the one (or a plurality of) logical links connected between the packet processing nodes to the maximum extent while allowing the mixed speeds to be mixed, and the information between the packet processing nodes is allocated. Bandwidth allocation for transfer is the sum of remaining bandwidth of one or more logical links existing between packet processing nodes. A structure capable of performing basis,
Allows information transfer with a transfer band larger than the maximum remaining band of the logical link existing between packet processing nodes,
It is characterized in that it is possible to increase the information transfer capability assigned to one certain logical link.

Further, in the above configuration, the information received from the plurality of routes may be rearranged in order and then the information may be transferred to the next node.

(Invention 12) In the present invention, in an information communication system configured by connecting a plurality of packet processing nodes and having multipath at the network level, each of the packet processing nodes performs multipath transfer or single path. It is characterized in that whether to pass or not is determined by using a predetermined conditional expression.

[0058]

[Action]

<1> (Invention 1) The network connection apparatus of the present invention (Claim 1) adopts the ATM protocol, which is a connection-type protocol, as a protocol between network connection apparatuses, and establishes and manages a connection between the network connection apparatuses. The means are provided in the network connection device.

As a result, the following operational effects are obtained.

(1) Since the communication control / management in the connected network and the connection between the network connection devices can be separately managed, the connection management closed in the unit network is unique in each unit network. Can be done.

(2) The network connection device of the present invention is a function under the control of the user of the unit network in which the network connection device is arranged, and the connection management between the network connection devices is the network management that operates the unit network. Since it can be carried out by a person, it is possible to realize an algorithm for connection between networks which is convenient for the operation of the network.

(3) Since the networks to be connected can be connected while having their respective network control functions, it is possible to clarify the functional division of the entire network composed of a plurality of networks.

(4) The network connection can be made without being aware of the protocol configuration of the connection destination network when making the network connection.

(5) When designing a network, the network connection method is not limited by the type of network, and the functional division of the entire network can be clarified, facilitating network design. It is possible to do.

(6) Since the communication path between the networks can be multiplexed and used by the connection identification,
The communication band in the communication path between networks can be effectively used.

(7) A communication service in a network that is performing connection-type high-speed communication such as ATM,
It can be provided across multiple networks.

(Invention 2) Since the network connection device of the present invention (claim 2) has a switching means with a high-speed hardware configuration such as an ATM switch in the network connection device, the following operational effects can be obtained. can get.

(1) Since it is possible to set a connection with a plurality of networks at the same time by using one network connection device, it is possible to realize a connection with a plurality of networks at the same time.

(2) A plurality of connections can be set between a plurality of networks, and the information sent through the connections is switched by a high-speed switching means having a hardware configuration, so that a high-speed operation across a plurality of networks can be achieved. The information transfer process can be realized.

(3) By selecting the network at high speed by the switching means in the network connection device, it is possible to increase the overall speed of protocol conversion in the network connection device.

(4) Since it is possible to simultaneously set a plurality of communication paths of the unique protocol between the network connecting devices between the connected networks, it is possible to flexibly set / change the communication band between the network connecting devices. As a result, it becomes possible to flexibly configure the connection form of the entire network by a plurality of networks.

(5) Since communication paths of a plurality of paths can be set for one network, the communication band in the inter-network communication path can be changed by dynamically changing the communication path between the networks when accepting a connection or operating a connection. Can be effectively used.

(6) Since device paths for network connection of a plurality of paths can be set for one network, even if a failure occurs in a network connection device or a communication path between network connection devices, it easily fails. By avoiding this, it is possible to provide a highly reliable network as a whole.

(Invention 3) In the network connection device of the present invention, one network is provided by arranging the protocol conversion means corresponding to the type of network to be connected for each one or a plurality of input / output ports of the network connection device. Information from a plurality of types of networks can be assigned to each connection in the communication path between network connecting devices corresponding to the destination address only by using the connecting device.

(Invention 4) A network connection device of the present invention is a network connection device using the network connection device according to the invention 2 (claim 2), and comprises a plurality of communication paths between connected networks. When it is possible, by providing the network connection device with failure detection / failure avoidance means for the communication path between network connection devices, the communication path set between networks can be used effectively. It will be possible to improve the reliability of the entire network.

(Invention 5) The network connection device of the present invention adopts the ATM method as the communication protocol between the network connection devices, so that the connection between the network connection devices is required when the end-to-end connection is set. Independently of the communication service quality, the connection setting means in the network connection device can set the communication service quality provided on the communication path between the network connection devices.

(Invention 6) The network connection device of the present invention adopts the ATM system as a communication protocol between the network connection devices so that connection-type communication and connectionless-type information are simultaneously sent from the network. When connecting between different networks, by providing different communication service qualities for connection-type information and connectionless-type information, communication quality services suitable for the type of information transferred between networks can be connected. Provide for connection between devices.

(Invention 7) The network connection device of the present invention is required at the time of setting the end-to-end connection when the connection type information is transferred between the network connection devices in the network connection device of the invention 5 or 6. By providing the communication service quality higher than the communication service quality to the connection between the network connection devices, the following effects can be obtained.

(1) It is possible to eliminate deterioration of the quality of discard service due to information transfer between network connection devices.

(2) It is possible to suppress the information delay time due to the information transfer between the network connection devices within the delay time expected by the connection management means in the connected network.

(3) In inter-network communication in which independent connection management is performed, even if the connection management means in each network independently perform communication service quality management, a discard service provided by the connection end-to-end It is possible to guarantee the quality.

(Invention 8) The network connection device of the present invention is required at the time of setting the end-to-end connection when the connection type information is transferred between the network connection devices in the network connection device of the above invention 5 or 6. According to the communication service quality and the resource information of the communication path between the network connection devices, a communication service quality lower than the requested communication service quality is provided to the connection between the network connection devices. By providing such a communication quality of service, although the quality of discard service due to information transfer between network connecting devices may deteriorate, it becomes possible to accommodate a large number of connections between network connecting devices at the same time. Therefore, the communication band between the network connection devices can be effectively used.

(Invention 9) The network connection device of the present invention is such that when the connectionless information is transferred between the network connection devices of the invention 5 or 6, the packet length of the connectionless information is transferred. To provide different communication service qualities between network connection devices according to characteristics such as burst length, burst length, source address of information, destination address of information, and type of application provided by connectionless information. Then, the following effects can be obtained.

(1) Even in connectionless type information, if the source address or the destination address is a terminal that manages the network or a terminal used in an emergency, that information By providing a high communication quality to the network, it becomes possible to transfer connectionless information between networks at high speed and with high quality.

(2) A network connection device that is assigned to long connectionless type information such as packet length and burst length when transmitting information between networks that are long distance apart By providing a high communication quality of service for the connection between the networks, it is possible to suppress the reduction of the effective throughput in the information transfer between the networks.

(3) A low communication service quality is provided for information that requires almost no real-time property, such as the current mail service, and high for information that requires real-time property, such as TV telephone information. By providing the communication service quality, it becomes possible to transfer information between networks suitable for the application provided by the information even in connectionless type information.

<2> On the other hand, in the information communication system of the tenth invention and the eleventh invention (claim 4), the information transfer capability of the packet processing node is allocated to the accommodation of the terminal and the connection between the packet processing nodes. The information transfer capability, which is divided into parts and is assigned to the connection between the packet processing nodes, is assigned to one or a plurality of physically different connection links between the packet processing nodes, and the one between the packet processing nodes is assigned. Or, a configuration in which bandwidth allocation for multiple physical links can be performed based on the total remaining bandwidth of inter-node connection links, and multi-link transfer at the data link level is allowed for information transfer between packet processing nodes Therefore, the remaining bandwidth of the connection links between individual packet processing nodes and the packet processing nodes Between the transfer bandwidth required for performing information transfer, max (remaining band inter-node connection link) <
Even when the relationship of bandwidth required for information transfer between nodes <Σ remaining bandwidth of inter-node connection links holds, the information transfer capacity of inter-node connection links does not become a throughput bottleneck, and high-speed packet transfer between nodes is possible. It is possible to realize a system in which the inter-node information transfer capability is less likely to become a bottleneck.

In the information communication system of the tenth and eleventh inventions (claim 4), the plurality of physical links between the packet processing nodes are allocated and routed as if they were one physical link. Therefore, the routing path can be simplified and the routing cost can be reduced.

Further, in the information communication system according to the tenth invention, the eleventh invention (claim 4) and the twelfth invention, the bandwidth allocation at the time of performing the information transfer between the packet processing nodes is one or more existing between the packet processing nodes. Based on the total remaining bandwidth of the physical links of a book, as if assigned to one physical link, and permitting information transfer with a transfer bandwidth larger than the maximum remaining bandwidth of the physical links existing between packet processing nodes. Therefore, it is possible to provide a system having high durability against high-speed and highly bursty information transfer.

Further, in the information communication system of the invention 10 and the invention 11 (claim 4), multipaths for the information transfer between the packet processing nodes are concentrated (logical at the data link level. The existence of multiple multipaths is allowed, but the logical multipath at the network level is not allowed.) Therefore, the number of hops of the multipath route becomes the same, and the problem of delay time variation due to the route, and It is possible to easily solve the problem of arrival order inversion due to delay variation when using multipath routes at the same time.

Further, in the information communication system of the tenth aspect of the invention, since the physical multipath at the network level is not allowed, the network level is lower than that at the network level. It has good consistency with the operation and management as a network without allowing logical multipath in the network, and it becomes possible to utilize the information transfer capability of the system components (nodes, links) of the network more efficiently. .

Further, in the information communication system of Invention 12,
In an information communication system that is configured by connecting multiple packet processing nodes and has multipaths inside the system, depending on the internal state of the network (number of nodes passing through, queuing state, etc.) when performing multipath transfer There is a difference between the case where multipath transfer is possible for high-speed transfer and the case where single path transfer is possible for high-speed transfer. Whichever transfer method is used, the network under multipath environment In order to solve the problem that the information transfer capacity of the system cannot be effectively utilized and to improve the throughput of the entire system configured by connecting the system components (nodes, links), the system components (nodes,
It is possible to provide a network that can effectively utilize the information transfer capability of (link).

[0093]

Embodiments of the present invention will now be described in detail with reference to the drawings.

<1> [Embodiment 1-1] FIG. 1 shows a basic embodiment of a network connection system using a network connection device according to the present invention. The network connection device according to the present embodiment (hereinafter, IWU (InterW
orking Unit)) and two different LANs
111 is connected to the LAN 112 and is a LAN
IWU221 to 111 and IWU222 to LAN112
IWU221 and IWU22
2 uses the ATM protocol as an original protocol between IWUs, and IWU 221 uses LAN 111 and A
IWU22 by performing protocol conversion with TM protocol
2 is configured to perform protocol conversion between the LAN 112 and ATM protocol. By using such a network connection method, it is possible to effectively use the band of the inter-IWU communication path 91 and provide a highly reliable communication path between the IWUs.

FIG. 2 shows an example of the internal schematic configuration of the network connection device IWU 221 of this embodiment.

In this IWU 221, the connected L
Information sent from AN111 LAN111
The information sent from the IWU communication channel is converted while the protocol is converted to the ATM protocol.
Inter-IWU protocol conversion means 31 for converting the TM protocol to the LAN 111 protocol is provided.
The inter-IWU protocol conversion means 31 is configured by using the destination information identification means 41 and the protocol conversion means 51, and the destination information identification means 41 identifies the destination information of the information sent from the LAN 111 and corresponds to the destination information. The IWU connection is allocated, and the protocol conversion unit 51 performs conversion of the frame format and the transmission rate. Along with this, the information sent from the IWU communication path is transferred from the ATM protocol to the LAN1.
In order to convert to the 11 protocol, the protocol converting means 51 converts the frame format and the transmission rate, and the destination information identifying means 41 from which connection in the IWU communication path the information came through.
Then, the processing of writing the destination information corresponding to the destination terminal in the LAN 111 is sequentially performed. The order of arrangement of the destination information identifying means 41 and the protocol converting means 51 is not limited to the order shown in FIG. 2, and the order may be reversed.

Further, in the IWU 221, there is a connection management means 71 for associating the destination information of the information sent from the terminal in the LAN 111 with the connection between the IWUs and managing and controlling the connection.
Information regarding the connection managed by the connection management means 71 is a destination / connection database (hereinafter abbreviated as destination / connection DB) 81 managed by the connection management means 81.
If the information regarding the connection is required, the information is read by the connection management means 71. Furthermore, the IWU management CPU (A1) is present in the IWU 221 in order to carry out management control of the entire IWU 221 such as operation and maintenance of each means provided in the IWU 221, monitoring of synchronization between IWUs, and data input / output control. To do. The IWU management CPU (A1) is connected to each means in the IWU via a communication path (not shown) having a bus structure, etc., and constantly monitors the means in each IWU during IWU operation.

The internal structure of the IWU 221 is not limited to the structure shown in FIG. 2, but the IWU does not have the connection management means 71 or the destination / connection DB 81 as in each embodiment described later. Alternatively, the connection management means 71 and the IWU management CPU (A1) may be operated on the same processor, or the IWU management CPU (A1) and the IWU internal means may be connected by a ring-shaped or star-shaped communication path. There may be cases such as when there is.

As described above, by performing the network connection as shown in FIG. 1 using the IWU shown in FIG. 2, the conventional network connection allows the network connection device adapted to the protocol of the destination network to be connected. Since it was necessary to use it, it would be sufficient to consider only the conversion between the operating network and the IWU original protocol (that is, ATM protocol),
It is possible to connect to a network without being aware of the type of network to which it is connected. Also, I
By adopting the connection-type ATM protocol as the WU-unique protocol, it is possible to provide a highly reliable communication path between IWUs and to effectively use the communication band of the IWU communication path. .

[Embodiment 1-2] FIG. 3 shows an embodiment of a network connection method for connecting networks using the network connection apparatus according to the present invention. This embodiment is an ATM network (ATM-LAN).
The network connection device of the present invention is applied to the case where such connection-type networks are connected to each other.

In this embodiment, the ATM-LAN 11
The connections in 5,116 are each ATM-LAN11.
5, 116 are managed and set by the connection setting means (CO in the figure) 441 and 442 respectively provided in the I / UUs 223 and 224, and the connection in the communication path between the IWUs 223 and 224 is the connection setting means 331 or 332 in the IWU.
Managed by. Further, in FIG. 3, ATM-L
Connections from the terminal 662 in the AN 115 to the terminal 664 in the ATM-LAN 116 CO-1, CO-4, C
Connections CO-2, CO-3, CO-5 set between the connection setting means for setting the connection between the O-6 and the ATM-LAN 115, 116.
Are also shown together.

FIG. 4 shows the IWUs 223 and 22 of this embodiment.
4 shows an example of an internal schematic configuration of No. 4. Data / connection selection means G2 which has received the information from the ATM-LAN
If the information sent from the destination information is connection setting request information, the information is sent to the connection setting means F2, and if it is normal Data, the information is sent to the ATM-LAN VP / VC identification means 42
Send to. The connection setting means F2 sends the connection setting request information and the resource management means E2.
From the resource management information sent from the server via the connection management means 72, it is determined whether the requested connection can be set up in the IWU communication path 92.
When the setting becomes possible, the information of the newly set connection is sent to the connection management means 72, and the connection management means 72 controls the ATM-LAN in-VP / VC identification means 42 in the IWU protocol conversion means 32. VP / VC connection in ATM-LAN115 and I
Match with the VP / VC connection in the communication path between WUs. The connection management means 72 stores information on the managed connection in the connection DB 82 to manage the connection.

An example of a procedure for setting a connection from the terminal 662 to the terminal 664 between the ATM-LANs 115 and 116 using the IWUs 223 and 224 will be described below. (1) The connection setting request from the terminal 662 is ATM
-Sent to the connection setting means 441 in the LAN 115.

The communication method between the connection setting means 441 and the terminal 662 may be an information transfer method by broadcasting or may be in accordance with a signaling procedure.
A method of setting the PVC in advance may be used. (2) Connection setting means 441 is ATM-LAN1
From the resource management information in 15, the terminal 662 and the IWU2
It is judged whether or not the requested connection can be established with the communication terminal 23. If the connection can be established, a connection setting request from the terminal 662 is sent to the IWU 223.

Here, the connection set by the connection management means 441 is the ATM-LAN in FIG.
This is the connection CO-1 in 115. Further, a connection CO-2 exists between the connection setting means 441 and the IWU223 internal connection setting means 331 (corresponding to F2 in FIG. 4). (3) The IWU 223 that has received the connection setting request information from the connection setting means 441 is the IWU 22.
If it is determined that the connection can be set between the communication terminals 3 and 224, the IWU connection CO-4 is set. (4) The connection setting means 331 which has set the requested connection between the IWUs 223 and 224 sends the connection setting request from the terminal 662 to the IWU224 internal connection setting means 332.

There is a connection CO-3 between the connection setting means between the connection setting means 331 in the IWU 223 and the connection setting means 332 in the IWU 224.

The connection CO-3 may be set according to the signaling procedure, or the connection may be set in advance. (5) Upon receiving the connection setting request, the connection setting means 332 in the IWU 224 sends the connection setting request to the connection setting means 44 in the ATM-LAN 116.
Send to 2.

Here, the connection setting means 442 and I
Connection setting means 332 in WU224 (F2 in FIG. 4)
(Corresponding to the above), the connection CO-5 exists.
The connection CO-5 may be set according to the signaling procedure, or the connection may be set in advance. (6) Connection setting means 442 is ATM-LAN1
IWU 224 and the terminal 6 from the resource management information in 16
It determines whether the requested connection can be set up during 64, and if the connection can be set up, IWU
A connection CO-6 is set between the H.224 and the terminal 664. (7) The same procedure is followed in reverse to send the connection connection completion information to the terminal 662, and the connection shown in FIG. 3 is completed.

The internal structure of the IWUs 223 and 224 is not limited to the method shown in FIG.
And the connection setting means 72 configured as one means (for example, the same CPU operates),
Various other construction methods are possible.

Further, the connection setting procedure is not limited to that shown here, and for example, the IWU 223 and the AT are not passed through the connection setting means 332 in the IWU 224.
The method of setting a connection between the connection setting means 442 in the M-LAN 116 to establish a connection between the terminals 662 and 664, and similarly without using the connection setting means 331 in the IWU 223, the ATM-LAN 1
A method of setting a connection between the connection setting means 441 in 15 and the connection setting means 332 in the IWU 224 to establish a connection between the terminals 662 and 664 is also conceivable.

[Embodiment 1-3] FIG. 5 is a network connection system for connecting a LAN (hereinafter, referred to as CL type LAN) for performing connectionless communication using the network connection device according to the present invention. An example will be shown.

In the network connection system shown in FIG. 5, I
A plurality of PVCs (Perm) are provided in the communication path between the WUs 225 and 226.
an anent virtual connection), CL type LAN1
13 and the data packet having the same destination information sent from the LAN 114 is transferred on the same PVC set between IWUs. Therefore, only data packets of the same destination information pass through the same PVC between IWUs, and it is possible to transfer data packets to the destinations for the set number of PVCs between IWUs at the same time.

By using the PVC of the ATM protocol as described above, even if the ATM protocol, which is a connection-type protocol, is used as an IWU-unique protocol, C
It becomes possible to provide communication between L-type LANs.

FIG. 6 shows the IWUs 225 and 2 shown in FIG.
An example of the internal schematic configuration of 26 is shown. In this IWU 225, there is an IWU protocol conversion means 33 for converting the information sent from the connected CL LAN 113 from the LAN 113 protocol to the ATM protocol. Here, the IWU protocol converting means 33 includes means 53 having the same function as the IWU protocol converting means 31 in the IWU 221 shown in FIG. However,
In the IWU 225, the destination information of the information transmitted from the terminal in the LAN 113 and the IWU 225, 22 are internally provided.
A PVC management means 73 for managing and controlling the PVC is provided in correspondence with the PVC set between 6 and 6. Also, as in the case of the IWU 221, the PVC management means 73
Information about PVCs managed by
Destination / PVC-managed by management means 73
It is stored in the DB 83. When information regarding PVC is required, the information is read by the PVC management means 73. Here, the PVC for setting the PVC between the IWUs when the IWU is installed or connected to the network.
Setting means B3 is present in the IWU 225, 226.

Further, in the IWU 225, the IWU 22
There is an IWU management CPU (A3) for performing operational control of each means in 5, IWU synchronization monitoring, and data I / O control management of the entire IWU 225. Although not shown in the figure, this IWU management CPU (A3)
Is connected to each means in the IWU 225 by a communication path such as a bus structure or a ring structure, and constantly monitors the means in each IWU 225 during IWU operation. However,
The internal configurations of the IWUs 225 and 226 are not limited to those shown in FIG. 6, but the IWU 22 shown in FIG.
Similar to the first embodiment, it can be realized in various configurations.

Since the PVCs shown in FIGS. 5 and 6 are set in advance and the communication between the IWUs is performed, the overhead for setting the connection between the IWUs can be reduced. It is possible to provide a communication path with a short transfer delay time. Also,
Since the PVC is set in advance, the table for connection identification required in the IWU and the management means for the table can be small, and the hardware amount of the IWU can be reduced.

[Embodiment 1-4] FIG. 7 shows an embodiment of a network connection system in the case of connecting a LAN performing connectionless communication using the network connection apparatus according to the present invention. In this embodiment, instead of using a fixed connection such as PVC as an ATM connection between network connection devices as shown in FIG. 5, a CL type L
When the CL type data packet arrives from the AN 113 to the IWU 227, the connection corresponding to the destination information is connected between the IWUs 227 and 228 on demand (On dema).
nd), and when a data packet having the same destination information arrives thereafter, the IWU communication is realized by a method of sequentially placing the data on the set connection and transferring the data. When it is determined that the connection is not needed, such as when the set connection is not used for a certain period of time, the connection is disconnected and the connection that corresponds to the destination information of the newly arriving data packet. Secure as.

Thus, the connection of the ATM protocol is set every time a CL type data packet arrives,
ATM, which is a connection-oriented protocol, releases the connection when it is no longer needed.
Even if the protocol is used, it becomes possible to provide the communication between the networks between the CL type LANs 113 and 114. FIG. 8 shows an example of the internal schematic configuration of the IWUs 227 and 228 shown in FIG. In the IWU 227, the connected C
Information sent from the L-type LAN 113 is sent to the LAN 113
IW to convert the existing protocol to the ATM protocol
As the U-to-U protocol conversion means 34, the IW shown in FIG.
Means 54 having the same function as the IWU protocol conversion means 31 in U221 are included.

Upon receiving the data packet from the CL type LAN, the destination information identifying means 44 sends the identified destination information to the On demand connection setting means C4.
On demand The connection setting means C4 has already determined that the destination is I according to the information from the connection management means 74.
It is determined whether or not the connection is set in the inter-WU communication path 94, and if it is already set as the connection, the set connection is assigned. If the destination information has not been set as a connection yet, a new unused connection is assigned to the destination information at the same time.
The assigned connection is used as the connection management means 7
4 to the destination information identifying unit 44, and further notifies the destination information and connection information newly associated with the destination information identifying unit 44. The notified destination information identifying means 44 converts the destination information into an IWU connection identifier, and then the IWU.
It will be sent to the inter-communication path 94.

Further, the connection management means 74 is an IWU.
The connection disconnection means D4 disconnects the connection when it is determined that the connection is not needed because the connection has not been used for a certain period of time or the transmission terminal has notified the end of communication. A command is sent, and the connection disconnecting means D4 issues a connection disconnecting command to the IWU protocol converting means 34. Further, the connection management means 74 uses the IWU 22.
Depending on the resource use status in the IWU communication path obtained by the resource management means E4 in 7, the On demand connection setting means C4 is instructed as to how many connections can be set later. Further, the connection management means 74 constantly monitors how many connections are set up in the IWU communication path 94, and the information of the connection is stored and managed in the destination / connection DB 84.

Further, also in the IWU 227, the IWU management CPU (A4) controls and manages each means in the IWU. Further, the IWU 227 is not limited to the configuration shown here, and naturally, the means allocating method shown in FIG. 2 can be considered.

On demand as shown in FIGS. 7 and 8.
By setting connections to different IWU communication paths, the number of connections set between IWUs can be freely changed without being limited by the number of PVCs set in advance. Also, since the overhead such as PVC setting does not occur at system startup,
It is possible to start up the system quickly. Further, since the requested communication bands are sequentially allocated in the communication path, the communication band in the IWU communication path can be used as much as necessary, so that the communication band can be effectively used.

[Embodiment 1-5] FIG. 9 shows an embodiment of a network connection system for connecting a connection type network using the network connection apparatus according to the present invention. Also in the network connection system according to the present embodiment, a configuration for connecting an ATM network (ATM-LAN) is shown. Further, in this embodiment, A
The connection in the TM-LAN 115 or 116 is ATM
-At the same time as being managed and set by the connection setting means (CO in the figure) 441, 442 in the LAN 115, 116, the connection in the communication path between the IWU 229, 230 is also set by the connection setting means 441 or 442 in the ATM-LAN 115 or 116. It will be managed and set.

In FIG. 9, the connection from the terminal 662 in the ATM-LAN 115 to the terminal 664 in the ATM-LAN 116 and the connection setting means for setting the connection between the ATM-LANs 115 and 116 are set. The connections that have been made are also listed.

FIG. 10 shows the IWU 229,2 shown in FIG.
An example of an internal schematic configuration of 30 is shown. The IWU protocol conversion means 35 which has received the information from the ATM-LAN,
ATM-under the control of the connection management means 75
The VP / VC connection in the LAN 115 and the VP / VC connection in the IWU communication path are matched. Since the IWUs 229 and 230 of this embodiment have no connection setting means, basically, it is only necessary to transfer the connection between the ATM-LAN and the IWU.

However, in this case, the resource management information in the IWU communication path needs to be sent to the connection setting means in the ATM-LAN 115 or 116 which is setting the connection in the IWU communication path. Therefore, in the IWU of this embodiment, IWU 229, 2
The connection management means 75 in 30 stores the resource management information in the IWU communication path from the resource management means E5,
ATM-LAN11 through resource information creating means B5
5 connection setting means 441 or ATM-LA
The data is sent to the connection setting means 442 in N116.

Also in this embodiment, the connection management means 75 stores the information of the managed connection in the connection DB 85 to manage the connection.
An example of a procedure for setting a connection from the terminal 662 to the terminal 664 between the ATM-LANs 115 and 116 as shown in FIG. 9 using the IWU 229 and 230 shown in FIG. 10 will be described below. (1) The connection setting request from the terminal 662 is ATM-
It is sent to the connection setting means 441 in the LAN 115. Here, the connection setting means 441 and the terminal 662.
The communication method between them may be an information transmission method by broadcasting, may follow a signaling procedure, or may be a method of setting a PVC in advance. (2) When the connection setting request is not directed to the terminal in the ATM-LAN 115, the connection setting means 441 is in the communication path between the resource management information in the ATM-LAN 115 and the IWU 229/230 transmitted from the IWU 229. From the resource management information of the terminal 662 and the IW
It is determined whether the requested connection between U229 and IWU 230 can be set up. If the connection can be set up, a connection setting request from terminal 662 is sent to IWU 229. Here, the connection management means 4
The connection set by 41 is the ATM in FIG.
A connection CO-2 in the LAN 115 and a connection CO-4 between the IWUs. Further, a connection CO- is provided between the connection setting means 441 and the IWU 229.
1 is present. The connection CO-1 may be set according to a signaling procedure or may be set beforehand. (3) Upon receiving the connection setting request information from the connection setting unit 441, the IWU 229 sends the connection setting request information to the IWU 230 by mounting the connection setting request information on the connection CO-3 for connection setting request information between IWUs. In addition, connection CO-3
The setting method may be in accordance with the signaling procedure, or may be a method of setting the connection in advance. (4) The IWU 230 sends the connection setting request information sent by CO-3 to the connection setting means 442 on the connection setting request information connection CO-5 in the ATM-LAN 116. (5) Connection setting means 442 is ATM-LAN11
From the resource management information in 6, the IWU 230 and the terminal 66
It judges whether the requested connection can be set up during 4 and if the connection can be set up, IWU2
The connection CO-6 is set between the communication terminal 28 and the terminal 664. (6) The same procedure is followed in reverse to send the connection connection completion information to the terminal 662, and the connection in FIG. 9 is completed.

IWU2 of this embodiment which provides the above-mentioned functions
The internal configurations of 29 and 230 are not limited to those shown in FIG. 10, and various configuration methods are conceivable as in the case of IWU 227 and 228 shown in FIGS. 7 and 8. Also, the connection setting procedure is not limited to the one shown here, and for example, I
Connection setting means 44 for setting connection between WUs
The connection setting means 442 instead of 1, or the AT to which the terminal that issued the connection setting request belongs
A method is conceivable in which the connection setting means for setting the M-LAN connection also sets the connection between IWUs.

By disposing the connection setting means in the unit network and in the IWU as in the configurations shown in FIGS. 3 and 4, when the connection setting load in the ATM-LAN is large, It is possible to connect to another network without concentrating the load of connection setting by connection on one connection setting means. Further, as in the configurations shown in FIGS. 9 and 10, the connection setting load between the IWUs is also borne by the connection setting means in the ATM-LAN, so that the CPU load in the IWU is reduced and the network connection is made. It is possible to reduce the load on the device.

By arranging the IWUs having both functions as described above in a well-balanced manner between the networks, it is possible to easily realize a well-balanced network configuration in which the connection setting load and the network management load are not biased. Like

[Embodiment 1-6] The respective embodiments have been described above. Next, the ATM-L using the network connection device according to the present invention described in FIGS. 11, 12 and 13 will be described.
Three examples will be shown in the case of establishing a network connection between ANs and realizing a copy connection across ATM-LANs.

Copy means 771, 7 shown in each figure
The copying means 72, 773, 774 includes a broadcasting means for transmitting the same information to all terminals to which the copying means performs the copy service, and information for some of the terminals to be serviced. A multicast means for copying and transmitting is possible. By providing the copying means with such broadcasting means and multicasting means, the information from the terminal 661 in the ATM-LAN 115 can be broadcast to all terminals in the ATM-LAN 116 by the following three configurations. It is also possible to send information to some specific terminals within the LAN 116. Below, the detail about each Example is shown.

(Embodiment 1-6-1) In the configuration shown in FIG. 11, the terminal 661 in the ATM-LAN 115 is the AT.
A copy connection is set to the terminals 662, 663 and 664 in the M-LAN 116.

From terminal 661 to terminals 662, 66.
As a method of setting a connection to 3,664, AT
The connection setting means 441 in the M-LAN 115 is A
The connection CO-1 in the TM-LAN 115 is set, the connection setting means in the IWU 231 or IWU 232 sets the connection CO-2 between the IWUs,
Connection setting means 442 in the ATM-LAN 116
Is a connection CO-3, C in the ATM-LAN 116
You may set O-4, CO-5, CO-6, CO-7. Alternatively, the connection setting means 441 in the ATM-LAN 115 has the connections CO-1 and CO-.
2 is set, and the connection setting means 442 in the ATM-LAN 116 sets the connections CO-3, CO-4, CO.
A method of setting -5, CO-6, and CO-7 may be used. Furthermore, the connection setting means 441 in the ATM-LAN 115 sets the connection CO-1, and the AT
The connection setting means 442 in the M-LAN 116 is connected to the connections CO-2, CO-3, CO-4, CO-5.
CO-6 and CO-7 may be set.

Connection CO-set in this way
1, CO-2, CO-3, CO-4, CO-5, CO-
6, CO-7 provides a copy function in the following procedure in the copy method in FIG. (1) The information sent from the terminal 661 is the connection C
O-1 and CO-2 from IWU231 to ATM-L
Sent to IWU 232 in AN 116. (2) The sent information is further transferred from the IWU 232 to the connection CO-3 via the exchange means 552 to the A
It is sent to the copy means 771 in the TM-LAN 116. (3) The information subjected to the copy processing by the copy means 771 is sent to the exchange means 552 again by the connection CO-4. (4) Exchange means 552 is connection CO-5, CO-
6, the terminal 6 in the ATM-LAN 116 by CO-7
Information is transmitted to 62, 663, and 664, respectively. Here, the copying means 771 may be provided inside the exchanging means 552, or the exchanging means 552 and the copying means 771 may be connected to the ATM.
-It may be configured as an independent unit within the LAN 116. Further, the copy means 771 is the ATM-LAN1.
In the case of an independent means within 16, the connection is set directly from the IWU 232 to the copy means 771, and the copy means 771 through the exchange means 552 to the terminal 66.
It does not matter if the connection is set to 2, 663 and 664. Further, the connection from the copying means 771 does not necessarily need to send the copied data to the exchanging means 552 which has sent the data to the copying means 771, and the connection from the copying means 771 is via another exchanging means in the ATM-LAN 116. The connection may be such that data is sent to the terminals 662, 663, and 664.

By using such a method, the AT
A copy connection from the M-LAN 115 to the ATM-LAN 116 can be provided. Further, by leaving the copy means to the copy means in the ATM-LAN 116, it becomes possible to provide the copy connection between the networks only by setting the connection of the band for one copy information as the connection between the IWUs. There is.

(Embodiment 1-6-2) Also in FIG.
The terminal 661 in the ATM-LAN 115 is an ATM-LAN.
A copy connection is set to the terminals 662, 663, 664 in the 116.

The copying method in this case has a structure in which a copying means is provided inside the network connection unit (IWU) so that a copying connection extending over the networks can be set. Also, the terminal 66
As a method of setting the connection from the terminal 1 to the terminals 662, 663, 664, ATM-L
Connection setting means 441 in AN115 is ATM-
Set the connection CO-1 in the LAN115, IW
The connection setting means in U233 or IWU234 sets the connection CO-2 between IWUs, and ATM-
Connection setting means 442 in LAN 116 is ATM
-Connections CO-3, CO-4 in LAN 116,
CO-5 may be set. Alternatively, ATM-
The connection setting means 441 in the LAN 115 sets the connections CO-1 and CO-2, and the ATM-LAN1
The connection setting means 442 in 16 connects the connection C
A method of setting O-3, CO-4, CO-5 may be used. Furthermore, the connection setting means 441 in the ATM-LAN 115 sets the connection CO-1,
The connection setting means 442 in the TM-LAN 116 is connected to the connections CO-2, CO-3, CO-4, CO-5.
May be set.

The connection CO-set in this way
1, CO-2, CO-3, CO-4, and CO-5 provide a copy function in the following procedure in the copy method in FIG. (1) The information sent from the terminal 661 is the connection C
O-1 and CO-2 from IWU233 to ATM-L
Sent to IWU 234 in AN 116. (2) The IWU 234 performs copy processing in the internal copy means 773. (3) Information that has been copied is sent from the IWU 234 to the terminal 6
Connection CO- set to 62, 663 and 664
3, CO-4, CO-5, terminals 662, 663,
Sent to 664.

By using such a method, as shown in FIG.
Instead of burdening the copying means in the ATM-LAN 116 with the copying means as in the case of 1, it is possible to provide the copying means inside the IWU 234.
It becomes possible to provide copy connections across networks without increasing the number of required connections in 116. Further, similar to the embodiment shown in FIG. 11, it is possible to provide a copy function across networks by only defining one connection between IWUs.

(Embodiment 1-6-2) Also in FIG.
The terminal 661 in the ATM-LAN 115 is an ATM-LAN.
A copy connection is set to the terminals 662, 663, 664 in the 116. Also in this case, FIG.
A network connection device (IW
The copy connection is provided in U) to realize a copy connection extending between networks. However, in the case of FIG.
Copy means 77 in the IWU 235 on the LAN 115 side
4 exists, but IWU23 on the ATM-LAN116 side
A method for realizing a copy means in the case where there is no copy means in 6 is considered. Terminal 661 in such a case
As a method of setting a connection from the terminal 662, 663, 664 to the ATM-LAN 11
The connection setting means 441 in 5 is ATM-LAN1
Connection CO-1 in 15 is set, and IWU235
Alternatively, the connection setting means in the IWU 236 is the IWU.
Connections CO-2, CO-3, CO-4 are set between the connection setting means 4 in the ATM-LAN 116.
42 is a connection CO- in the ATM-LAN 116
5, CO-6, CO-7 may be set. Alternatively, the connection setting means 441 in the ATM-LAN 115 has the connections CO-1, CO-2, CO-3,
CO-4 is set, and the connection setting means 442 in the ATM-LAN 116 sets the connections CO-5 and CO-.
6, CO-7 may be set. Furthermore, the connection setting means 4 in the ATM-LAN 115
41 sets connection CO-1, ATM-LAN
The connection setting means 442 in 116 is connected to the connections CO-2, CO-3, CO-4, CO-5, CO-6, 6.
CO-7 may be set.

Connection CO-set in this way
1, CO-2, CO-3, CO-4, CO-5, CO-
6, CO-7, the copying means is provided by the following procedure in the method in FIG. (1) The information sent from the terminal 661 is the connection CO
-1 to IWU 235. (2) Information copy processing is performed by the copy means 774 inside the IWU 235. (3) The copied information is the connection CO-2, C
It is sent to IWU 236 by O-3 and CO-4. (4) IWU 236 connection between IWUs CO-
2, CO-3, CO-4 and the ATM-LAN 116 internal connections CO-5, CO-6, CO-7 are matched, and the respective connections are connected. (5) The copy information sent from the IWU 236 is sent to the terminal 66 by the connections CO-5, CO-6 and CO-7.
2,663,664.

By using such a method, the IW
Even when one IWU does not have a copy means when performing U-to-U communication, it is possible to provide a copy means in a connection across networks without imposing a burden on the copy means that may exist in the ATM-LAN 116. It will be possible. Further, similarly to the advantage of the embodiment shown in FIG. 12, the copy connection from the ATM-LAN 115 can be provided to the ATM-LAN 116 even when there is no copy means in the ATM-LAN 116.

As in the case of the method shown in FIG. 13, when a copy function is provided in the transmission side network, information is copied by the copy means provided in the ATM-LAN 115 as shown in FIG. After going, I
By sending information from the terminal 661 to the terminals 662, 663 and 664 via the WUs 235 and 236, it is possible to provide a copy means that spans networks.

[Embodiment 1-7] Next, FIG. 14 and FIG.
FIG. 5 shows an example of a network configuration in the case of using the above-mentioned network connection device according to the present invention to provide datagram communication across ATM-LAN. Here, CLSF (Connection-Less Service) is used as a means for providing datagram communication between ATM and LAN.
Function) processing means are used in two embodiments.

Example 1-7-1 In FIG. 14,
The terminal 662 in the ATM-LAN 115 is the ATM-LA.
A connection is set to the terminal 663 in N116 so as to perform datagram communication.

Here, as a method of setting a connection from the terminal 662 to the terminal 663, the ATM-LAN 115 is used.
The connection setting means 441 in the ATM-LAN11
5, the connection CO-1 in IWU 237 or IWU 238 is set by the connection setting means in IWU 237 or IWU 238, and the ATM-LAN 11
The connection setting means 442 in 6 is ATM-LAN1
The connection CO-3 in 16 may be set. Alternatively, the connection setting means 441 in the ATM-LAN 115 sets the connections CO-1 and CO-2,
Connection setting means 442 in the ATM-LAN 116
May set the connection CO-3. Furthermore, the connection setting means 441 in the ATM-LAN 115 sets the connection CO-1, and the AT
The connection setting means 442 in the M-LAN 116 may set the connections CO-2,3.

Further, instead of setting each connection by the connection setting means, the connection is previously assigned to the terminal in the ATM-LAN and the IWU or the CLSF processing means in the IWU by a setting method such as PVC. When the terminal 662 performs datagram communication, information may be transferred to the preset connection and transferred. Alternatively, the connection between IWUs is not set by the connection setting means, but P
A method of setting a connection such as VC may be used. Furthermore, an on-demand connection setting method that sets a connection each time datagram information arrives may be used.

Connection CO-set in this way
1, CO-2, CO-3 provide datagram communication in the method of FIG. 14 by the following procedure. (1) Information sent from the terminal 662 is the connection CO
-1 by CLSF processing means 881 in IWU 237
Sent to. (2) In the CLSF processing means 881, from the destination information of the datagram information, the destination terminal is the ATM-LAN 11
5 does not exist, the information is transmitted to the IWU 238 by the connection CO-2 in the IWU communication path. (3) The IWU 238 reads from the destination information of the transmitted datagram information which terminal in the ATM-LAN 116 the information is. (4) From the IWU 238 to the terminal 663, the datagram information is transmitted to the terminal 663 by the connection CO-3 which is newly set or preset.

Here, it is the IW that provides the CLSF function.
CLSF processing means 881 in U237 may be used, or IW
The CLSF processing means 882 in U238 may be used.

(Embodiment 1-7-2) Also in FIG.
The terminal 662 in the ATM-LAN 115 is an ATM-LAN.
A connection is set so that the terminal 663 in 116 can perform datagram communication.

Here, as a method of setting a connection from the terminal 662 to the terminal 663, the ATM-LAN 115 is used.
The connection setting means 441 in the ATM-LAN11
Set connection CO-1, 2 in 5 and IWU23
9 or the connection setting means in IWU240 is IW
Set the connection CO-3 between U and ATM-LAN
Connection setting means 442 in 116 is ATM-LA
Connections CO-4 and 5 in N116 may be set. Alternatively, the connection setting means 441 in the ATM-LAN 115 causes the connections CO-1, 2, 3
May be set, and the connection setting means 442 in the ATM-LAN 116 sets the connections CO-4 and CO-5. Furthermore, the connection setting means 441 in the ATM-LAN 115 is connected to the connection CO-1,
2, the connection setting means 442 in the ATM-LAN 116 may set the connections CO-3, 4, and 5.

Further, each connection is not set by the connection setting means, but between the terminal in the ATM-LAN and the CLSF processing means or between the CLSF processing means and the IW.
A connection is assigned in advance between U by a setting method such as PVC, and when the terminal 662 performs datagram communication, the information is transferred to the preset datagram connection. You may do it. Similarly to the embodiment shown in FIG. 14, the connection CO-3 between IWUs may not be set by the connection setting means, but may be a method of setting a connection such as PVC in advance. Alternatively, an on-demand connection setting method that sets a connection each time datagram information arrives may be used.

Connection CO-set in this way
1, CO-2, CO-3, CO-4, and CO-5 provide datagram communication by the following procedure in the method of FIG. (1) Information sent from the terminal 662 is the connection CO
-1 to the CLSF processing means 883 in the ATM-LAN 115. (2) If the CLSF processing means 883 determines from the destination information of the datagram information that the data is not the delivery data into the ATM-LAN 115, the IWU is established by the connection CO-2 established with the IWU 239.
Send information to 239. (3) The IWU 239 transmits the information to the IWU 240 by the connection CO-3, which indicates whether the connection is newly set between the IWUs or is preset based on the destination information of the transmitted datagram information. (4) In the IWU 240, when the sent information passes through the connection for datagram information (CO-3 etc.), the information is transferred to the C in the ATM-LAN 116.
It is transmitted to the LSF processing means 884. (5) The CLSF processing means 884 reads from the destination information of the sent datagram information which terminal in the ATM-LAN 116 the information is, and sends the information to the terminal 663 by the connection CO-5. To do.

By adopting the configuration shown in FIG. 14 or FIG. 15, the CLS in the ATM-LAN is realized.
It is possible to provide the datagram communication across the ATM-LAN when the ATM-LAN without the F processing means is connected or when the CLSF processing means is provided as the means in the ATM-LAN. become able to do.

[Embodiment 1-8] Next, FIG. 16 shows an embodiment of a network connection system for performing connection-type communication across a plurality of networks using the network connection device according to the present invention described above. Indicates. In the figure, from the terminal 662 in the ATM-LAN 115 to the ATM.
-Connections CO-1, CO-2, CO- through the LAN 117 to the terminal 663 in the ATM-LAN 116
3, CO-4, CO-5, connection setting means 441, 442, 44 for setting the connection
Connection CO for connection setting request preset between 3, 333, 334, 335, 336
-6, CO-7, CO-8, CO-9, CO-10, C
O-11 is shown.

Although the connection network is assumed to be a connection-type network here, the connection network is not limited to the connection-type network, and the various connectionless-type networks described above are used. A connectionless network can be connected by using the network connection method.

Terminal 662 to terminal 663 using FIG.
The procedure for setting up a connection to is shown below. (1) The connection setting request signal from the terminal 662 is AT
It is sent to the connection setting means 441 in the M-LAN 115. Here, the connection setting means 441 and the terminal 6
The communication method between the 62 may be an information transmission method by broadcasting, may follow a signaling procedure, or may be a method of setting a PVC in advance. (2) The connection setting means 441 sets the connection CO-1 between the terminal 662 and the IWU 241 because the connection request terminal (terminal 663) is not in the ATM-LAN 115, and the connection setting request information is set to the connection CO-6. Is sent to the connection setting means 333 in the IWU 241. However, if the setting cannot be made, a connection refusal signal is sent to the terminal 662. (3) Connection setting means 333 is IWU 241 and IW
If a connection can be set between U245, connection CO-2 is set, and connection setting request information is sent to the connection setting means 335 in IWU245 using connection CO-7. However, when the connection cannot be set, a connection refusal signal is sent to the connection setting means 441 using the connection CO-6. (4) The connection setting means 335 sends the connection setting request information to the ATM-LAN using the connection CO-8.
117 to the connection setting means 443. (5) Since the terminal (terminal 663) of the connection connection request destination is not in the ATM-LAN 117, the connection setting means 443 connects the IWUs 245 and 246 with the connection CO.
-3 is set, and the connection setting request information is sent to the connection setting means 336 in the IWU 246 using the connection CO-9. However, if it cannot be set, a connection refusal signal is sent to the connection setting means 335.

Since the ATM-LAN 117 is a network connecting the ATM-LAN as a backbone network, the connection setting means 443 determines which AT is the terminal to which the connection is requested.
It knows whether it belongs to the M-LAN. (6) The connection setting means 336 connects the IWU 246 and I
If a connection can be set between WU242, connection CO-4 is set, and connection setting request information is sent to connection setting means 334 in IWU242 using connection CO-10. However, when the connection cannot be set, a connection refusal signal is sent to the connection setting means 44 by using the connection CO-9.
Send to 3. (7) The connection setting means 334 sends the connection setting request information to the connection setting means 442 in the ATM-LAN 116 using the connection CO-11. (8) Since the terminal of the connection setting request destination is in the ATM-LAN 116, the connection setting means 442 sets I
Connection CO-5 is set between WU242 and terminal 663, and connection connection information is set to connection CO-11.
Connection setting means 334 in the IWU 242
Send to. However, if it cannot be set, a connection refusal signal is sent to the connection setting means 334. (9) The connection connection information is sequentially sent back, and finally the connection is established between the terminals 662 and 663.

Here, for the sake of simplicity, the IWU is also provided with a connection setting means, but in this case as well, the connection between the IWUs is made by the ATM-LAN as in the previous case.
The connection setting means 441, for example, the connection CO-2 is set by the connection setting means 441.
It does not matter if the configuration is set by. Further, the connection setting means in the IWU on the connection setting request side sets the connection between the IWUs, and in this case also, the connection setting means of the IWU on the opposite side of the connection setting request side sets the connection between the IWUs. For example, the connection CO-2 is connected to the connection connecting means 333.
The configuration may be made by the connection setting means 335 instead of by the setting.

Further, in FIG. 16, the connection is set in advance between the connection setting means in the IWU and the connection setting request information is transferred, but the connection setting request information is transferred. A configuration in which connection setting request information is transferred using a newly set connection, for example, IWU
After setting the connection CO-2 between the 241 and the IWU 245, the connection setting request signal is sent by using the connection CO-2.
It does not matter even if it is configured to transfer to 5.

By using such a network connection system, it is possible to construct a LAN spread over a wide area by connecting a plurality of ATM-LANs and various networks by a high-speed backbone ATM network. In addition, with the configuration shown in FIG. 16, the ATM-LAN1
Even when 15 and the ATM-LAN 116 are not directly connected to each other between LANs, a high-speed ATM connection can be set between the terminal 662 in the ATM-LAN 115 and the terminal 663 in the ATM-LAN 116. It is possible to realize a high-speed network configuration using a plurality of ATM-LANs.

Also, the backbone network A
The connection setting means 443 in the TM-LAN 117 performs all connection management between IWUs, so that the ATM-LAN operated by the user is connected to the backbone network.
If one IWU is installed and connected to the ATM-LAN 117 without adding new means inside, other ATM-LA
It is also possible to communicate with N, and the network operation and network expansion are made easier than ever before.

[Embodiment 2-1] FIG. 17 shows an embodiment of a network connection system using the network connection apparatus according to the present invention. In FIG. 17, the LAN (A1
11, A112, A113) and network connection devices A221, A222, A2 provided in respective LANs.
23 shows an example of a network configuration when connected by 23. Here, each IWU (A221, A222,
A223) includes protocol conversion means A441, A44.
2, A443, and connection setting means A881,
A882 and A883 are provided respectively.

FIG. 18 shows the IWU (A2 shown in FIG.
21-A223) shows a schematic diagram of an example of the internal configuration.

This IWU has, as interface means on the LAN side, modularized protocol conversion means A31-1 to A31-n obtained by modularizing the protocol conversion means between the connected LAN and ATM protocol. Here, n is the number of interfaces on the LAN side that the IWU can have. Also, it is expected that a buffer means for holding packets during processing time required for protocol conversion and information collision will be required even in this case, but this buffer means can be provided in the modularized protocol conversion means. And the buffer means itself is I
It is also possible to configure the modularized protocol conversion means to have an interface with the buffer provided in the WU main body. Also, the modularized protocol conversion means A3
1-1 to A31-n convert the address system of the connected LAN and the address system of the inter-IWU protocol (that is, ATM protocol), and connect the intra-LAN protocol of the connected LAN and the IWU communication path. It is configured to have an interface for inputting / outputting connection management information.

Further, in the present IWU, together with the connection setting means AM1 for setting the ATM connection between the network connecting devices, the connection set between the network connecting devices, the address information, the connection information and the like are managed. Connection management means A
There is 71. The connection management means A71 is configured to store and manage the addresses of the LAN and IWU connections managed therein and information regarding the connections in the connection database A81 existing in the IWU. Further, the resource management means AE1 for obtaining information required for connection setting from each IWU communication path is provided.

In this resource management means AE1, IWU
As the resource information of the inter-communication path, the resource information such as the number of connection settings, the used bandwidth (peak value, average value, etc.) and the communication bandwidth of each communication path is managed. The resource management information collected by the resource management means AE1 is sent to the connection management means A71 to be used as information for connection management, and a connection for setting the connection in the IWU communication paths A91-1 to A91-n. It is also sent to the setting means AM1 and used as resource information in the communication path necessary for connection setting.

In this IWU, the modularized protocol conversion means A31-1 to A31-n are used.
In order to have a removable structure, the module connection means A is provided between the IWU body and the modularized protocol conversion means.
H1-1 to AH1-n are present. This provides means such as a connector for attaching / detaching the modularized protocol converting means to / from the IWU main body, and the means of the IWU main body is configured to be once terminated by this module connecting means. Even if there is no modularized IWU protocol conversion means ahead of the above, the IWU main body can provide that means.

Further, the module connecting means AH1-1 to AH1-1.
Between the AH1-n and the switching means AZ1, there are I / F means AJ1-3 to AJ1-n for matching the input interface to the switching means AZ1. Further, the switching means AZ1 to the IWU communication path A91
There are AJ2-1 to AJ2-n for matching the output interface to -1 to A91-n. These I
The / F means is an I / F means AJ3-1 in FIG.
~ AJ3-n and AJ4-1 to AJ4-n provide the same functions.

Here, the protocol converting means between the IWU and the network is modularized so that it can be attached to and detached from the IWU main body. The protocol conversion means may be embedded in the IWU body in advance. Alternatively, only the protocol conversion means is I
The protocol conversion means and the IWU main body may be installed separately from the WU main body so that information can be transferred by some means. In addition, IWU in IWU
There is an IWU management CPU (AA1) for performing overall timing control and management of means to be provided by the IWU. Although the IWU management CPU exists independently in the figure, an internal bus or the like exists between the IWU management CPU and each means inside the IWU of FIG. 18 in the actual IWU.

By implementing such an IWU configuration, the selection of the transmission destination communication channel is made, as compared with the case where the transmission destination communication channel of the received information is determined by software processing as in the conventional multi-protocol router. Since it is possible to perform high-speed hardware processing, it is possible to provide high-speed inter-network communication even when making network connections such as connecting multiple networks. Further, the protocol conversion means in the IWU is modularized so that it can be attached / detached to / from the IWU main body, so that it is possible to easily provide a plurality of patterns of network connection using one IWU and to connect to the network. Even when the configuration is changed or the communication capacity between networks and the number of networks to be connected themselves are expanded, it is possible to easily provide the means.

[Embodiment 2-2] Next, FIG. 19 shows a plurality of ATM-Ls using the network connection device according to the present invention.
An example of a network connection method for connecting an AN will be described.

In FIG. 19, ATM-LAN (A1
14, A115, A116) for each ATM-L
Network connection device A224, A225, A in AN
22 shows an example of a network configuration in the case of connection by 226. Here, each IWU (A224, A22
5, A226), the connection setting means A884,
Although the configuration is such that A885 and A886 are provided, this is not necessarily the case, and an IWU having no connection setting means may be used.

FIG. 20 shows the IWU (A22 shown in FIG.
4 to A226) is a schematic view of an example of the internal configuration of the present invention.
As in the IWU shown in FIG. 18, the IWU shown in FIG. 20 includes a switching unit AZ2 for switching the connection in the communication path connected by the IWU.

This switching means AZ2 basically provides the same function as that of the switching means AZ1 shown in FIG. 18, performs self-routing switching by hardware such as ATM switching means, and operates in the IWU. Information exchange at high speed and IW
It is possible to set a connection for performing high-speed information transfer in the communication path connected by U, and to enable high-speed information transfer processing between networks.

The number of input / output interfaces of the switching means AZ2 is set according to the information transfer amount (throughput) set between networks and the number of networks simultaneously connected by the IWU. However, it is also possible to set a large number of interfaces of the switching means AZ2 in advance and provide a new communication path to the IWU by using the redundant interface when the network is expanded or the network configuration is changed. Is.

Further, the ATM-L is included in the IWU of this embodiment.
As the interface means on the AN side, Data / connection selection means AK for sorting the transmitted packet into a connection setting request packet and an information packet.
2-1 to AK2-n exist.

If the transmitted packet is a connection setting request packet, the connection setting request packet is sent to the connection setting means AM2 in the IWU. This connection setting means AM2
It corresponds to the connection setting means A884 to A886 in FIG. 19 and has a function of setting the connection between IWUs. When the transmitted packet is a normal information packet, it is directly transmitted to the VP / VC converting means AL2-1 to AL2-n. The VP / VC conversion means performs address conversion between the connected ATM-LAN and the IWU-specific protocol, and the input VP / VC and output VP notified from the connection management means.
The VP / VC conversion is performed according to the / VC conversion table. Here, n is an ATM that IWU can have
-It is the number of interfaces on the LAN side, and normally it is possible to have the same number of modules as the number of input / output interfaces of the switching means AZ2 in the IWU, but it is not necessarily the same number of modules as the number of interfaces of the switching means AZ2. Not the ATM-LAN side or IWU
The number of interfaces according to the usage pattern of the inter-communication path is IW
It is possible for U to have it.

In order to perform such VP / VC conversion,
In the IWU of this embodiment, there is a connection management means A72 for managing the connection between the ATM-LAN side and the IWU. Connection management means A
Reference numeral 72 indicates the address of the ATM-LAN or IWU connection managed there and information regarding the connection.
It is configured to be stored and managed in the connection database A82 existing in the WU. Also, each IW
Resource management means AE2 for obtaining resource information from the U-to-U communication paths A92-1 to A92-n exists in the present IWU.

In this resource management means AE2, IWU
As the resource information of the inter-communication path, the resource information such as the number of connection settings, the used bandwidth (peak value, average value, etc.) and the communication bandwidth of each communication path is managed. The resource management information collected by the resource management means AE2 is sent to the connection management means A72 to be used as information for connection management, and a connection for setting connections in the IWU communication paths A92-1 to A92-n. It is also sent to the setting means AM2 and used as resource information in the communication path necessary for connection setting.

Furthermore, the VP / VC conversion means AL2-1 to AL2-1
Between the AL2-n and the switching means AZ2, I for adjusting the input interface to the switching means.
/ F means AJ3-1 to AJ3-n are present. This I /
The F means is the IWU sent from the VP / VC conversion means.
Switching means AZ according to the destination information of the information packet according to the inter-protocol (that is, ATM protocol).
2) Processing such as conversion into a physical address (routing address) by a self-routing algorithm inside 2 or conversion into a packet format within the switching means AZ2 (parallel expansion or addition of a routing header) is performed. Also, the output of the switching means AZ2 is IWU
I / O for sending into the inter-communication path A92-1 to A92-n.
An I / F for converting the packet converted into the format in the switching means by the F means AJ3-1 to AJ3-n into the format of the IWU original protocol again
The means AJ4-1 to AJ4-n are the switching means AZ.
2 and the IWU communication paths A92-1 to A92-n.

ATM-connected in this way
The information from the LAN is configured to be sent to a predetermined IWU communication path by the switching means after receiving the VP / VC conversion. Further, it is expected that a buffer means for holding a packet during the processing time required for VP / VC conversion or information collision will be required in the IWU, but this buffer means should be provided in the VP / VC conversion means. It is also possible to have it in the I / F means.

In the IWU of this embodiment, there is also an IWU management CPU (AA2) for controlling the timing of the entire IWU and managing the functions to be provided by the IWU. In the figure, the IWU management CPU (AA2) exists independently, but inside the actual IWU, the IWU management C
An internal bus or the like exists between the PU and each unit inside the IWU in FIG.

By implementing such an IWU configuration, when connecting a plurality of ATM-LANs, ATM
-Since it is possible to freely connect between LANs and set up an ATM connection across networks, it is possible to realize high-speed information transfer between terminals belonging to different ATM-LANs. It will be possible.
Further, since the number of interfaces provided by the IWU can be set arbitrarily, the network configuration can be made flexible, and the IWU can be used even when the network is expanded or the network configuration is changed.
It is possible to easily provide the IWU corresponding to the network form without significantly changing the means. Further, using this IWU, some of the interfaces input to the switching means are not provided with the Data / connection selection means and the VP / VC conversion means, and the ATM-LA is used.
The VP / VC in N is input to the I / F means as it is and mapped to the physical address in the switching means, and some of the output interfaces of the switching means corresponding to the mapped addresses are directly used in the ATM-L.
By allocating to the communication path in AN and allocating the remaining several to the communication path between IWUs, a part of the means of IWU is ATM-
While being used as a normal switching node in a LAN, it is possible to provide a network connection means between ATM and LAN by the same IWU.

Further, the IWU shown in FIGS. 17 and 18,
The IWUs shown in FIG. 19 and FIG. 20 do not have to be used independently, and when connecting a connection type network (ATM-LAN etc.) and a connectionless type network, etc. 19 and 20 are installed in the network, the IWUs shown in FIGS. 17 and 18 are installed in the connectionless network, and the IWUs are connected to each other by using the ATM protocol as an inter-IWU original protocol. It is possible to

Further, each I in FIG. 17 and FIG.
In the communication path between WUs, it is possible to set the connection by using the VP / VC value of each unique address system. In this case, the connection management means A71, A72 shown in FIGS. 18 and 20 determines which IWU is connected to which IWU communication path by the connection D.
B (A81, A82) stores and manages. further,
The connection management means A71 and A72 are provided for each communication path in order to switch connection identifiers of different address systems sent from the IWU communication path with one switch, and the I / F means AJ2-1 to AJ2-1. n,
In the AJ4-1 to A-n, a switch internal routing table is provided for each address system of the IWU communication path. I / F means AJ2-1 to AJ4-1
.. through n, information having the same VP / VC value is sent from different IWU communication paths by switching the information sent from each IWU communication path according to the routing table for each address system. However, as the routing information in the switch, those V
P / VC values can be treated as independent.

The following effects can be expected by setting the VP / VC values of the different address systems for each IWU communication path. (1) Even when the number of connection identifiers required for one network connection measure increases, each connected network has an independent address system, so it can be identified by one network connection device. It is possible to increase the number of connections that can be made. (2) Since the number of connection identifiers is the same as the number of addresses with which information may be transferred between network connection devices, the number of bits for connection identification can be reduced. It is possible to simplify the address conversion process in the above, and to reduce the storage area for storing the address installed in the network connection device. (3) Since address management between network connection devices will be performed independently between each network connection device,
Connection setting between network connection devices When setting a fixed connection for signal transfer or a fixed connection for information transfer, be aware of the connection set with other network connection devices each time you set the connection. Without doing so, it is possible to configure such that even if the connections are set independently, the connections will not be mixed.

[Embodiment 2-3] Next, referring to FIG. 21, a plurality of connectionless type networks using the network connection device according to the present invention (see the above-mentioned Embodiment 2-1 and Embodiment 2-2). An example of a network connection method when a connection spanning between is set is shown.

In FIG. 21, a CL type LAN (A1
11) The server A661 in the ATM-LAN (A
117), connections CO-1, CO-2, CO-3 set for transferring information to the server A663 in the CL type LAN (A112), and the CL type LAN
CL type LA directly from server A662 in (A111)
The connections CO-4 and CO-5 set to transfer information to the server A664 in N (A112) are shown.

CL type LAN (A111), backbone ATM-LAN (A117) and CL type LAN here.
(A112) and ATM-LAN (A117) connections CO-1 and CO-3, and CL type LAN (A11)
Connection CO between 1) and CL LAN (A112)
-4 is basically the same as the procedure of connection setting in the embodiment 1-
3 (FIG. 5) and the connection establishment method shown in Embodiment 1-4 (FIG. 7). Further, the procedure for setting the connection of the connection CO-2 in the backbone ATM-LAN (A117) can be basically performed by the same procedure as the connection establishing method shown in the embodiment 1-8 (FIG. 16).

Since such a network connection method can be realized by using the IWU of this embodiment, CL
The terminals in the type LAN (A111) are CL type LAN (A1
12) As a method of setting the connection when communicating with the terminal in (a) ATM-LAN: CL 117 via CL-LAN:
Method to set up connection with terminal in A112 (b) CL type LAN: A111 directly to CL type LAN: A
It is possible to use two types of connection selection, which is a method of setting a connection for the terminal in 112.

Therefore, the CL type LAN (A111) and CL
It is possible to reduce the bottleneck of the network connection between the type LAN (A112) and to provide the network selection by the high-speed switching means of the hardware configuration even when the connection with a plurality of networks is made. Therefore, it becomes possible to realize a faster network connection than the network connection method for selecting a network by the conventional software processing.

Further, in FIG. 21, for example, ATM-L
AN (A117) falls into a failure state and IWU (A23
0) through ATM-LAN (A117) to IWU
Even if the connection connected to (A231) becomes unavailable and the communication between the server A661 and the server A663 cannot be performed, the IWU (A230) and the IWU can be used in the network connection method according to the present embodiment.
Connection C is newly added in the communication path with (A231).
O-5 is set, and until now the connection CO-1 or CO
It is possible to secure the communication from the server A661 to the server A663 by dynamically changing the information carried in -3 to be carried in the connection CO-5.

As described above, by using the IWU of the present invention, it is possible to provide a failure avoiding means in case of a network failure even in an inter-network communication path.

In this case, the following two methods can be considered as the method of providing the failure avoidance means. (a) The IWU management CPU in the IWU provides a failure avoidance means for the IWU communication path by the management means in the IWU, and autonomously selects the path of the IWU communication path or changes the path in the IWU. Method (b) The failure avoidance means for the IWU communication path is left to the failure avoidance means in the network to which the IWU belongs, and the processing such as route selection and route change of the IWU communication path is also externally managed. The IWU receives the result only and changes the route setting in the IWU.

By using the above method, I
It is possible to increase the reliability of the entire network without changing the reliability of the WU communication path itself. In a network that provides a means as a backbone network of a plurality of CL-type LANs such as ATM-LAN (A117), the IWU in each CL-type LAN is always a plurality of IWUs in the backbone ATM-LAN.
21. For example, in FIG. 21, IWU (A
230) connects to each IWU (A227, A228),
IWU (A231) is each IWU (A228, A229)
It can be configured to connect to.

By doing this, the backbone ATM
-There will be no CL-type LAN that makes the backbone ATM-LAN unusable due to the failure of one IWU in the LAN. That is, even if the IWU (A228) fails, both CL type LANs (A111, A112) are connected to the backbone ATM-LAN (A117).

In the network connection system of this configuration, the reliability of the network as a whole can be increased by adopting such a network connection configuration.

[Embodiment 2-4] Next, FIG. 22 shows another embodiment of the network connection system in the case where a connection is set up between a plurality of connection type networks using the network connection apparatus according to the present invention. Indicates.

In FIG. 22, ATM-LAN (A1
14) From the terminal A661 in the backbone ATM-L
ATM-LAN (A115) via AN (A117)
Connection CO-1 set to the terminal A668 in the
CO-2, CO-3, CO-4, CO-5 and ATM-
ATM-directly from the terminal A663 in the LAN (A114)
The connections CO-6, CO-7, and CO-8 set in the terminal A666 in the LAN (A115) are shown.
Furthermore, IWU (A2 in ATM-LAN (A114)
35), as described in the description of FIG. 20, the IWU according to the present invention can also provide a function as a switch node in the ATM-LAN.
2, the connection CO-9 via the IWU (A235) can be set between the terminal A664 and the terminal A665.

The connection setting procedure here can be basically performed by the same procedure as the connection establishing method shown in the embodiment 1-8 (FIG. 16). Also, as in the previous case, the connections CO-2, CO in the IWU communication path
-4 and CO-7 may be performed by the connection setting means A771, A772, A773 in the ATM-LAN, or the connection setting means existing in the IWU is a connection between the IWUs, which is not shown in the figure. May be set.

Such a network connection method can be realized by using the IWU according to the present invention.
The terminal in the TM-LAN (A114) is an ATM-LAN.
When performing communication with the terminal in (A115), as a connection setting method, (a) a connection with the terminal in the ATM-LAN (A115) is set via the backbone ATM-LAN (A117). Method (b) ATM-LAN: ATM-LAN directly from A114
It is possible to use two types of connection selection, which is a method of setting a connection for the terminal in (A115),
ATM-LAN (A114) and ATM-LAN (A11
It is possible to reduce the bottleneck of network connection with 5) and to increase the number of routing patterns, which allows flexible connection settings and ATM-LA.
It is possible to realize the connection setting using the band effectively as a whole, without using the communication band in N in a biased manner.

Further, even when connecting to a plurality of networks, the network selection can be provided by the high-speed switching means of the hardware configuration. Therefore, the network connection method for selecting the network by the conventional software processing. It is possible to realize a faster network connection.

Further, in FIG. 22, for example, ATM-L
AN (A117) falls into a failure state and IWU (A23
5) from the IWU via the ATM-LAN (A117)
Even if the connection that was connected to (A236) becomes unavailable and communication between terminal A661 and terminal A668 is no longer possible, the network connection method of this configuration allows the connection between IWU (A235) and IWU (A236). Connection CO-10 is newly set in the communication path of
-2, dynamically changing the connection CO-4 and the connection CO-5 to be replaced, so that the connection CO-1 and the connection CO-10, and the connection CO-10 and the connection CO-5 are replaced. Thus, it is possible to secure the communication from the terminal A661 to the terminal A668.

As described above, by using the IWU according to the present invention, it is possible to provide a failure avoiding means at the time of network failure even in the inter-network communication path.

In this case, as the method of providing the failure avoidance means, the following two methods can be considered as in the case of connecting the connectionless network described in the previous embodiment (FIG. 21). (a) The IWU management CPU in the IWU provides a failure avoidance means for the IWU communication path by the management means in the IWU, and autonomously selects the path of the IWU communication path or changes the path in the IWU. Method (b) The failure avoidance means for the IWU communication path is left to the failure avoidance means in the network to which the IWU belongs, and the processing such as route selection and route change of the IWU communication path is also externally managed. IWU receives only the result and changes the route setting in the IWU. By using the above method, the reliability of the IWU communication channel is not changed and the reliability of the entire network is improved. It is possible to increase the degree. Also, ATM-LAN
In a network providing a means as a backbone LAN of a plurality of ATM-LANs such as (A117), the IWU in each ATM-LAN must be the backbone A.
A configuration for connecting to a plurality of IWUs in TM-LAN,
For example, in FIG. 22, IWU (A235) is IWU (A2
32) and IWU (A233), and IWU (A23
6) can be configured to be connected to the IWU (A233) and IWU (A234).

By doing this, the backbone ATM
There will be no ATM-LAN that will be unable to use the backbone ATM-LAN due to the failure of one IWU in the LAN. That is, even if the IWU (A233) fails, the ATM-LAN (A114) and the ATM-LA
Both N (A115) and backbone ATM-LA
It is connected to N (A117).

In the network connection system of this configuration, the reliability of the network as a whole can be increased by adopting such a network connection configuration.

[Embodiment 2-5] FIG. 23 shows I of the present invention.
An example of a network connection method in the case of using WU to provide datagram communication across a plurality of networks is shown.

In FIG. 23, the ATM-LAN (A1
14) from the terminal A662 within the ATM-LAN (A11
7) via the terminal A6 in the ATM-LAN (A115)
63-ATM-LA when sending datagram information to
The connection set in N is shown.

The connection setting procedure can be basically performed by the same procedure as the connection establishing method shown in the embodiment 1-7 (FIGS. 14 and 15). ATM-LAN (A
114) is a method for implementing datagram information communication in FIG.
In the case of FIG. 15, the datagram communication realizing method in the ATM-LAN (A115) corresponds to the case of FIG. In addition, in the ATM-LAN (A117), IW
Both Us (A237, A238) have CLSF processing means inside, but as in the previous case, this may also have CLSF processing means inside the IWU, and ATM-LAN. A method of setting a connection in advance between the CLSF processing means in (A117) and the IWU (A237, A238) may be used.

Further, although not shown in FIG. 23, IW
A connection is set in the communication path between the U (A239) and the IWU (A240), and the CLS in the IWU (A239) is set.
F processing means and CLSF in ATM-LAN (A115)
By setting a connection with the processing means, the terminal A
It is also possible to provide datagram communication from 662 to terminal A 663.

[Embodiment 2-6] FIG. 24 shows an embodiment of a network connection system in the case where a plurality of communication paths are set between networks using the IWU according to the present invention.

In FIG. 24, the ATM-LAN (A1
14) and the ATM-LAN (A115) (that is, between the IWU 241 and the IWU 242) are connected by one IWU communication path, and the ATM-LAN (A115) and the ATM-LAN (A115) are connected.
Between LAN (A116) (ie IWU242 and I
WU243) is connected by three IWU communication channels,
TM-LAN (A114) and ATM-LAN (A11
6) (that is, between the IWU 243 and the IWU 241) is connected by two IWU communication paths.

Since the communication band of each IWU communication path is usually the same, the ATM-LAN (A115) and AT
ATM-LAN between M-LAN (A116)
This means that a communication capacity three times as large as the communication path between (A114) and the ATM-LAN (A115) is provided.

By using such a network connection method, it is possible to provide a larger communication band between the ATM and LAN than when the ATM and LAN are connected by one communication path, and if necessary. ATM-LAN
It is possible to provide a network connection method that can select the communication band between them.

At a certain point in time, as shown in FIG.
M-LAN (A115) and ATM-LAN (A116)
Communication capacity was the most necessary between the ATM-LAN (A) and the ATM-LAN (A) due to the increase in the number of communication terminals or the communication capacity per terminal.
115) to ATM-LAN (A114) and ATM-LAN (A115) to ATM-LAN (A115)
If the communication capacity to 116) becomes almost the same, the ATM-LAN (A115) to the ATM-LAN
A method for newly installing an IWU communication path to (A114), and ATM-LAN (A115) and ATM-LAN in FIG.
One of the three IWU communication paths set up with the LAN (A116) is connected to the ATM-LAN (A115).
It is possible to easily provide the required communication band between the networks at the time of operating the network by performing processing such as replacement with the communication path between the ATM-LAN (A114) and the ATM-LAN (A114).

[Embodiment 2-7] FIG. 25 shows a schematic configuration example of a main part of an IWU when the IWU according to the present invention is used to provide various functions such as a CLSF function and a copy function in the IWU. .

As described above, the ATM-L shown in FIG.
When the terminal A662 in the AN (A114) sends out a datagram type information packet, the connection CO-1 set in the ATM-LAN causes the IWU.
It is adapted to be sent to the CLSF processing means in (A235). When this is applied to FIG. 25, the connection CO-1 from the terminal A662 in FIG.
I / F means AJ5-i (i = 1 to 1) in the IWU inside
From n), it is set to the CLSF means AP3 through the switching means AZ3.

Although not shown in the figure, in this case IW
The connection management unit in U identifies the connection to be connected to the CLSF processing means in advance, and
The routing table for determining the routing in the switching means is set so that all the information packets sent to the processing means are sent to the output port of the switching means connected to the CLSF processing means.

Also, the connection CO-2 in FIG.
Is an appropriate IWU communication path A93 from the CLSF processing means AP3 in FIG. 25 through the switching means AZ3 again.
-I (i = 1 to n) is set.

By the same method, it is possible to provide the copy function between IWUs by the copy means AQ3. However, the copying means can be supplied only by the method shown here, and the following methods can be considered. (a) A method for providing a copy means by providing a means for sending the same information packet a plurality of times inside the switch (b) A certain I / F means AJ5 in the preceding stage of the switching means
-I (i = 1 to n) provides a copy means by sending the same information packet a plurality of times (c) I / F means AJ6 after the switching means
-A method of providing a copy means by sending the same information packet a plurality of times by i (i = 1 to n) (d) When using a multistage switch as the switching means,
A method of providing a copy means for the entire multi-stage switch by sequentially copying the information by the switches of each stage, and here, as a means to provide in the IWU, CLSF is provided.
Although only the processing means and the copying means are shown, by using the switching means inside the IWU as shown in FIG. 25, the IWU management CPU, the connection setting means, the connection disconnecting means, the connection managing means, the resource managing means, etc. It is possible to provide the means of selecting.

This function can also be provided by arranging the means for providing each function before the input of the switching means AZ3 or after the output of the switching means AZ3.

[Embodiment 3-1] FIG. 26 shows an embodiment of the network connection method in the case of connecting a plurality of connectionless networks using the network connection apparatus of the present invention.

In FIG. 26, three IWUs (A24
4, A245, A246), seven CL-type LANs
(A121 to A127) are connected, and three IWUs (A244, A245, A246) are connected.
Connections connecting between CO-1, CO-2, CO-
3 and connections CO-4, CO-5, CO-6 for connecting the connection setting means (A887, A888, A889) respectively provided in each IWU.

In such a configuration, IWU (A24
4) CL type LAN (A121, A122) and IWU
A protocol conversion means for inter-protocol (that is, ATM protocol) is installed, and CL type LA is installed in IWU245.
A protocol conversion unit between the N (A123, A124, A125) and the IWU protocol is installed, and the IWU246
A protocol conversion unit between the CL type LAN (A126, A127) and the inter-IWU protocol is installed in. IWU
The functions provided by (A244, A245, A246) are
Basically, the function is the same as that of the means shown in the embodiment 2-1 (FIG. 18), but in this embodiment, the protocol converting means to be provided is installed in a plurality of types of IWU main body. . In the figure, three IWUs (A244, A24
5, A246), a connection CO for information transfer
Although only one of each of -1, CO-2, and CO-3 is shown, in reality, as many connections as the number of connections according to the information transfer amount can be set.

Further, according to the present embodiment, according to the amount of information required between IWUs, as shown in the embodiment 2-6 (FIG. 24), I
By changing the number of IWU communication paths installed between WUs,
It has a configuration capable of providing a communication path adapted to the amount of information between IWUs.

Further, as a connection setting method in the IWU of this configuration, a connection setting between IWUs when connecting between CL type LANs as described in Embodiments 1-3 and 4 (FIGS. 5 and 7) and the like. The method can be used.

By using such a network connection system, due to the flexibility of the ATM system in terms of communication speed, a plurality of CL LANs having different speeds can be connected using the same IWU. Can be realized.

For example, it is the communication speed of Ethernet.
Five 10Mbps networks and a FDDI communication speed of 100Mbps network are accommodated in the same IWU, and the communication speed of ATM communication method between IWUs is 155Mb.
It is possible to configure so that IWU communication can be performed by using ps.

For the same reason, it is possible to reduce the increase in the number of communication paths between networks when the number of networks increases, which is a problem in the method of arranging IWU for each network. In addition, it has a configuration that can solve the problem of a decrease in the throughput of the router when the number of connected networks increases, which is a problem when connecting to a network using a multi-protocol router. Further, in FIG. 26, C
Although a configuration is shown in which a plurality of L-type LANs are connected to the IWU and accommodated, the network to be connected as described above is not necessarily limited to the connectionless type. For example, which CL in FIG. Even if the type LAN is replaced with a connection type network,
WU can be used to provide similar network connections.

[Embodiment 3-2] In FIG. 27, the respective network connection devices as shown in FIGS. 18 and 26 are used.
ATM-LAN which is a connection type network
An example of a network connection method for connecting a plurality of LANs (such as Ethernet) having low-speed interfaces is shown in FIG.

In FIG. 27, the low speed interface L
The ANs (A121, A122, A123) are concentrated by the IWU (A248) having the configuration shown in FIG. 26, and the IWU (A24) having the configuration shown in FIG.
It is configured to be connected to the ATM-LAN (A117) via 7). In addition, ATM-LAN (A1
17) Terminal A662 is a low-speed interface LAN
The connection set up when communicating with is shown. In this embodiment, the IWU (A248) functions as a multiplexer for connecting a so-called low speed LAN to the ATM-LAN.

Also, the terminal A662 in the ATM-LAN (A117) is connected to the IWU (A
247). IWU (A247) and IW
Connection CO for communication with U (A248)
-1 and the connection CO-3 for connection setting request information transfer are set. A low-speed LA is provided at the end of each modularized protocol conversion means in the IWU (A248).
N is connected, IWU (A247) and IWU (A24
8) ATM-LAN by IWU communication path with
Communication with terminals in (A117) is possible.

Here, the method of concentrating the low-speed LAN is not limited to the method shown in FIG. 27, and a plurality of communication paths may be set between IWUs instead of only one. Also, without using IWU (A248), IWU
(A248) may exist in the ATM-LAN (A117) and directly condense the low-speed LAN. Further, when the low-speed LAN is a connection type LAN, a connection may be set between the IWU (A247) and the terminal in the low-speed LAN.

By using the low-speed LAN concentrating system by the network connecting device of the present invention as described above, it is possible to use the conventional network connecting device instead of the ATM-LAN.
(A117) and low-speed LAN (A121, A122, A1
23), it is necessary to perform network connection processing with each network, and after a protocol conversion between networks by one network connection device, network connection can be performed using multiplexed high-speed communication paths. This makes it possible to easily accommodate a low speed LAN in a high speed network (for example, ATM-LAN).

Also, one high-speed LAN (for example, ATM-
Since a plurality of low-speed LAN interfaces can be accommodated at the end of the (LAN) side interface, it is possible to realize network connection effectively using the network connection interface. Also in FIG. 27,
Similar to the case of FIG. 26 described above, a configuration is shown in which a plurality of CL LANs are connected and accommodated in one IWU, but the network to be connected as described above is not necessarily connectionless. However, the IWU of this embodiment can provide the same network connection regardless of which CL type LAN in the figure is replaced by the connection type network. ing.

[Embodiment 4] Next, FIG. 28 shows an embodiment of a network connection method when a plurality of networks are connected using the network connection apparatus according to the present invention. In FIG. 28, in the situation where a plurality of connection-type or connectionless-type networks are connected,
Terminal A661 and ATM in ATM-LAN (A114)
-A terminal A662 in the LAN (A115) establishes a connection using the connections CO-1, CO-2, and CO-3 to perform communication.

When the IWU of the present invention is used, for example, when a failure occurs in the IWU communication path in which the connection CO-2 is set, IWU (A252) or IWU (A
The failure detecting means in 253) can detect a failure in the IWU communication path.

In this embodiment, the AT is provided on the communication path between the IWUs.
Since M connections are set and communication is performed, it is conceivable to detect a failure by using the OAM function which is a failure detecting means of ATM. Here, the IWU (A252) or IWU (A253) that has detected the failure changes the connection to the connection CO-2 to the IWU (A2).
52) and IWU (A253). IWU (A252) and IWU (A2
If another inter-IWU communication path exists between the connection unit 53 and 53) and a connection having the same band as the connection CO-2 can be set in the communication path, the connection CO- in FIG.
A new connection such as No. 4 is set and a change is made so that it is used instead of the connection CO-2.

Also, IWU (A252) and IWU (A2
53), if a new connection cannot be set up in the communication path with the
A new communication path between U (A252) and IWU (A253) is searched. In FIG. 28, the connection CO-5 when connecting the IWU (A252) and the IWU (A253) via the IWU (A254) is shown.

As a failure of IWU itself, IWU (A
253), the failure of the IWU itself is detected by the failure detection means in the IWU (A253) or the IWU (A25).
3) will be detected by the failure detecting means in the IWU (A252, A254) connected. Further, a method of detecting a failure by a failure detecting means in the network to which the IWU (A253) belongs can be considered.

As a method for avoiding the failure of the IWU (A253) in which the failure is detected, the internal function (component) of the IWU (A253) is duplicated or a plurality of IWUs are installed in the ATM-LAN (A115). Connect between networks,
If a failure is detected, the method is switched to the internal function (component) of the IWU (A253) standby system, or the ATM-LA.
A method of connecting to a network using another IWU in N (A115) is possible.

As in the case of the failure detection method, the IWU switching instruction and the internal function switching instruction for avoiding such a failure are issued by the IWU (A253) itself and connected by the IWU (A253). IWU (A25
2, A254), etc., or IWU (A2
It is conceivable that the failure management means in the ATM-LAN (A115) to which 53) belongs belongs.

FIG. 29 shows a conceptual diagram of one embodiment of the internal configuration of the network connection apparatus according to this embodiment. In FIG. 29, the configuration diagram is shown focusing on the failure avoidance means of the IWU, so that the internal configuration parts of this IWU are not all those shown in FIG. 29, and naturally the other components shown in FIG. 18 and FIG. Means will be included.

The IWU shown in FIG. 29 includes OAM information management means AW5 and OA for detecting a failure in the communication path between IWUs.
There are M cell add / drop means AV5-1 to AVn. O
The AM information management means AW5 sends a failure detection packet (OAM cell) to the IWU communication path periodically or at an arbitrary timing in order to detect a failure in the IWU communication path.

[0248] Further, the OAM cell flowing in the IWU communication path periodically or at an arbitrary timing is received, and O
OAM in inter-IWU channels collected by AM cells
Based on the information, it monitors whether there is any failure in the IWU communication path. Further, the OAM information management means AW5 is an OA sent from the OAM information management means in another IWU.
OAM information and OAM that I know about M cell
IWU of the transmission source with OAM information indicating whether the IWU in which the information management means AW4 is installed is normal
Processing such as sending it back to the OAM information management means inside is performed.

Next, the OAM cell drop / insert means AV5-1.
.. through n, O sent from the OAM information management means AW5 in the connection set up in the IWU communication path.
Processing is carried out in which AM cells are transferred and transferred, or conversely, only OAM cells are selected from the transferred cells in the connection in the IWU communication path and transmitted to the OAM information management means AW5. Currently, standardization of OAM cell identification methods is progressing with CCITT and the like, but OAM cell identification methods do not necessarily have to follow CCITT specifications, and the IWU of the present invention needs to comply with standardization specifications such as connection between LANs. OA when used in an environment without
Various methods can be considered for identifying the M cell. Method using GFC field in header area of ATM cell, V
Method using part of PI / VCI field, PT or C
Method using LP field, method using HEC area, method using bit in AAL header trailer,
Alternatively, a method using an information field may be considered.

Here, referring to FIG. 28 and FIG. 29, I
An example of a failure avoidance method when a failure of the inter-WU communication path is detected is shown below. 1) IWU in which the connection CO-2 is set by the OAM cell transmitted from the OAM information management means AW5 in the IWU (A252) or IWU (A253)
Consider a case where a failure (such as link disconnection) in the inter-communication path is detected. Here, the failure is detected and IWU (A252)
The OAM cell notified to the
It may be a cell. 2) IWU (A2) that received the OAM cell that detected the failure
52), OAM cell add / drop insertion means AV5-1 to AVn
The OAM cell is sent to the OAM information management means AW5 in the IWU (A252) to notify the failure detection. 3) When the failure information is sent to the OAM information management means AW5, the OAM information management means AW5 communicates with other IWUs connected to the IWU (A253) to which the communication path in which the failure is found is connected. Determine if a road exists. 4) If another communication path exists, the connection management means A75 is asked whether or not the same communication band as the connection CO-2 can be set in that communication path. In addition, IWU (A25
The configuration may be such that the connection management unit A75 also determines whether or not there is another communication path between 2) and the IWU (A253). 5) When a new connection can be set between the IWU (A252) and the IWU (A253), the connection management means A75 newly adds the connection CO-4 to the connection setting means AM5 instead of the connection CO-2.
Is set between IWU (A252) and IWU (A253). 6) Connection C by connection setting means AM5
O-4 is set so that communication between the terminals A661 and A662 can be performed in place of the connection CO-2 which cannot be used due to a failure. Heretofore, from terminal A661 to connection CO-1, CO-2, CO-
The information that has been transferred to the terminal A 662 through 3 is transferred through the connections CO-1, CO-4, and CO-3. 7) If a new connection cannot be established between the IWU (A252) and the IWU (A253), the connection management means A75 issues a connection setup request to the connection setup means AM5 to set up another route. . 8) IWU (A2
52) through IWU (A254) to IWU (A25
The connection CO-5 up to 3) is set so that the communication between the terminals A661 and A662 can be performed in place of the connection CO-2 which cannot be used due to the failure. Here, the information that has been transferred to the terminal A662 from the terminal A661 through the connections CO-1, CO-2, and CO-3 until now is the connection CO-1 and CO-.
5, will be transferred through CO-3. Note that C
The connection setting method of O-5 may be a method of sequentially transferring connection setting requests to the IWU, a method of setting the connection by the IWU knowing the network connection configuration in advance, or the IWU.
(A252) I connecting the connection setting request
A method of broadcasting to all WUs may be used. 9) If a new connection in place of connection CO-2 cannot be set between IWU (A252) and IWU (A253), either give up the connection setting for failure avoidance or reduce the bandwidth as the connection. Therefore, a failure avoidance method such as securing only the connection by using the same method as the third to eighth processing is conceivable.

By providing the IWU with the failure avoiding means shown in FIGS. 28 and 29, when a failure occurs in the IWU communication path, the IWU itself searches for another communication path to avoid the failure. This makes it possible to provide a failure detection / avoidance function between IWUs without imposing a new burden on the connected network.

In the network connection method using the IWU of the present invention, since a large number of IWU communication paths can be set, the IWU communication paths are installed with redundancy when making network connections. 17 and 1
The reliability can be improved even in the normal network connection method such as No. 9, but by providing the IWU with the failure avoidance means such as the OAM means, the more reliable network connection method is provided. Things are possible.

Also, the OAM management means AW shown in FIG.
IWU by monitoring the failure status of its own means
It is also possible to detect its own failure. It is also possible for the OAM management means AW5 in FIG. 29 to detect the failure of another IWU by monitoring the failure status of the connected IWU. Further, when the OAM management means detects a failure of itself or a failure of another IWU, the OAM management means sends the above-mentioned system switching command at the time of failure detection to the IWU which has detected the failure. You can do it.

As described above, the OAM management means inside the IWU can detect not only the inter-IWU communication path but also the failure of the IWU itself and can avoid the failure, so that higher reliability can be obtained. It is possible to provide a network connection method.

[Embodiment 5] Next, FIG. 30 shows the IW of the present invention.
FIG. 1 is a schematic diagram of an embodiment of the internal means configuration of the IWU that enables the management of the communication quality of service (hereinafter abbreviated as QOS) related to information transfer using the U even in the IWU communication path. Show.

Since FIG. 30 focuses on the means for managing communication quality, only a part of the IWU internal means is shown. Although not shown in the figure, naturally, the means inside the IWU shown in FIGS. 18 and 20 are present.

In FIG. 30, the required QOS of each connection required at the time of connection setting is the QOS management means AR in the IWU through the connection management means A76.
6 will be notified. The QOS managing means AR6 stores the required communication quality information in the QOS information database AS6, and the QOS guaranteeing means AT6-1 to AT6- installed for each IWU communication path originating from this IWU.
Notify n.

The QOS assurance means AT6-1 to AT6-n monitor the information in the communication path on a connection-by-connection basis and monitor whether the information flowing in each connection satisfies the requested QOS and is sent out. If there is a packet that violates the request, the following processing may be performed. (a) Discard the violating packet. (b) The violating packet is made to wait until it is in a state commensurate with the requested communication quality. (c) Raise a flag that it is in violation, and preferentially discard it at the time of congestion. (d) Notify the user of the violation. (e) Notify the I / F means and the VP / VC conversion means before the switching means in the IWU that there is a violation, and make the information transmission wait a little. With such a method, it is possible to secure the required QOS and perform communication even in the IWU communication path.

Among the functions provided by the QOS guaranteeing means shown here, for the functions such as (a) and (c) above, a method using a policing function performed in ATM communication can be considered. Similarly, among the functions provided by the QOS guaranteeing means, in the processing for the violating packet,
The function such as (b) may be provided by using the shaping function performed by ATM communication. Note that a sliding window method, a leaky bucket method, or the like can be considered as a method of policing or shaping.

Further, since the QOS guaranteeing means provides a function very similar to the policing function and the shaping function in this way, it is provided by one function module integrating these functions in the network connection device of the present invention. By doing so, it is possible to easily provide a policing function and a shaping function between networks.

Here, the method of providing the communication quality of service (QOS) provided by the communication service guaranteeing means is not limited to the method of guaranteeing the same QOS required at the time of setting the connection even in the connection in the IWU communication path. However, the following various methods are conceivable as the providing method. (Method 1) IW a different QOS from the connected LAN
Method provided by U-to-U communication path This method 1 is an IW when connecting an ATM-LAN.
In the U-to-U communication path, instead of setting the same QOS as that set at the time of connection setting, for example, IW
By providing only one type of QOS in the U-to-U communication path, a QOS of higher quality or a lower quality of QOS than the requested QOS is provided.
Communication between IWUs will be carried out by the OS, but IWU
In this method, the function for providing QOS is reduced. Further, between IWUs, the QO of the peak rate (maximum required quality of service) of the communication quality always required
By allocating S, the use efficiency of the communication band in the IWU communication channel decreases, but it provides a high-quality communication channel between IWUs, eliminates the deterioration of QOS due to network connection, and provides QOS in IWU. It is also possible to consider a method that makes it possible to reduce the function. (Method 2) QO in advance on the PVC in the IWU communication path
Method of setting S In this method 2, when PVCs are set between IWUs to provide communication between IWUs, QOs are preliminarily assigned to each PVC.
S is set and the bandwidth and Q of the requested connection are set.
There is a possibility that waste of bandwidth and QOS will be inefficient or insufficient by sequentially allocating PVC suitable for the set of OS, but IWU
This is a method that can reduce both the connection setting function and the QOS providing function that are provided internally. Also, I
When assigning a PVC connection to the WU communication path,
By allocating the bandwidth of the peak rate required by the connection, QO within the IWU as in the previous case
A method that can reduce the S providing function can be considered. (Method 3) Method for Providing QOS without Bandwidth Allocation (Only Connection is Set) This method 3 is particularly applicable to connections that do not require real-time transfer delay time in communication such as datagram communication. Sets a connection on the IWU communication path, but does not allocate a band to the connection, and provides a QOS that can transfer information of the connection only when there is a free band on the IWU communication path. Is.

By using this method, the connection of the voice communication and the real-time image information communication which requires the real-time property is robbed of the band by the non-real-time connection such as the datagram communication. As a result, the IWU It is possible to avoid the effect that it takes time to transfer information between them and the real-time property is impaired. Also, with this method, even if the communication band of the IWU communication channel is limited, the band in the communication channel is not robbed by receiving a large number of datagram communication connections, and the connection of the real-time connection is established. It is possible to increase the reception probability. In addition, the QOS provided by this method is
A QOS that can be applied as one pattern of the QOS in the case of the method 2 in which S is assigned and in the case of the method 1 in which a QOS different from the ATM-LAN is provided between IWUs. is there. (Method 4) For each multiple connections (for example, in VP units)
Method of setting QOS In this method 4, it is not possible to monitor or guarantee QOS for all connections set between IWUs.
This is a method of collectively managing the QOSs of a plurality of connections set between IWUs, since U will be given a heavy load. That is, for example, VP in ATM system
Q using a connection identifier such as the relationship between
A method is such that the OS management is performed only for each VP, and the QOS of the VC included in the VP is not managed. In addition, the types of QOS that can be set between IWUs are limited in advance, VPs are assigned to the set QOSs, and the connection requested by the terminal is an appropriate VP (set between IWUs) for each QOS. (Corresponding to QOS), and the monitoring and guarantee of QOS are performed in units of the VP, thereby reducing the overhead of monitoring for all connections.

Even in such a QOS providing method, it is possible to set the "no bandwidth allocation" QOS shown in the above method 3 as one pattern of the QOS. (Method 5) Method of providing QOS by performing retransmission control between IWUs This method 5 is a method of guaranteeing QOS between IWUs, and a method of guaranteeing at the exit of IWUs as described above, It is not a QOS guarantee method by providing functions such as policing and shaping at the interface point of the IWU to the IWU communication channel by using a method such as providing retransmission control means at the data link layer level in the communication channel , A QOS guarantee method by retransmission control between IWUs is conceivable.

[Embodiment 6] Next, in FIG. 31, using the network connection device according to the present invention, between networks (hereinafter abbreviated as CO & CL mixed LAN) in which connection type and connectionless type communication are mixed. 1) shows an example of a network connection method for connecting a network.

Here, a mixed CO & CL LAN (B11
1) from the terminal B661 in each IWU (B221, B22
2) to the terminal B663 in the CO & CL mixed type LAN (B112), and the connections CO-1, CO-2, C
O-3 is set and connection-type communication is performed. Further, from the terminal B662 in the CO & CL mixed type LAN (B111) through each IWU (B221, B222), the terminal B in the CO & CL mixed type LAN (B112).
664, connections CO-4, CO-5, CO-
6, CO-7 and CO-8 are set, and connectionless communication is performed.

Here, as a connectionless type information transfer method, it is considered that CLSF processing means is used to provide connectionless type communication in an ATM environment. However, the connectionless type information providing method is The method is not necessarily limited to this method, and may be the case in which a connection-type information area and a connection-type information area are secured in a frame and communication is performed, as in FDDI-II. A method of preparing a slot for transferring connection-type information and a slot for transferring connection-type information in advance such as a reservation method may be used. Also, the connections CO-2 and CO-6 between IWUs are the same IW.
It may be set in the U-to-U communication path, or may be set in different IWU-to-IUU communication paths.

In this case, the IWU of the present invention is connected to the connection CO-2 in which the connection in the IWU communication path is performing connection-type communication and the connection CO-6 in which connection-less communication is performed. Therefore, different QOS will be provided.

FIG. 32 shows a conceptual diagram of an internal configuration example of the IWU of this embodiment. Since FIG. 32 focuses particularly on the means for providing the QOS of FIG. 30, only the means for providing the QOS is shown in the figure. QOS guaranteeing means AT6-1 to ATn in FIG. 30 correspond to B11 and B12 in FIG. Naturally, FIG.
20 and the internal means of the IWU as shown in FIG. FIG. 32 shows a QOS providing method in the case where one type of QOS is provided for each IWU connection for transferring connection type information and each IWU connection for transferring connectionless type information.

Hereinafter, referring to FIGS. 31 and 32, QO will be described.
An example of a method of guaranteeing cell discard quality in S will be shown. Figure 32
The QOS guarantee method is the QOS guarantee means A in FIG.
T6-1 to n have such a configuration that they are arranged in front of the I / F means AJ11-1 to n, and the switching means AZ.
QO in case of guaranteeing QOS before input to 6
The S guarantee method is shown. (1) The information sent from the terminal B661 or CLSFB991 is input to the QOS assurance means B12 before being input to the switching means BZ2 in the IWU. (2) In the information identifying means B22 in the QOS guaranteeing means B12, the same information as in the previous case is supplied to the connection managing means B7.
It will be sent from 2. (3) The information identifying means B22 is a waiting buffer B for each QOS that should be guaranteed for each connection according to the information.
The information is sent to 33 and B34. (4) If connection-type information is transferred to the waiting buffer B33 and connectionless-type information is transferred to the waiting buffer B34, QOS is applied to the connection between IWUs by the algorithm for extracting the information from the waiting buffers B33 and B34. Will be guaranteed. Information is preferentially extracted from a queuing buffer that guarantees QOS with high cell discard quality, and the switching means performs ACK control to wait for information transmission in the preceding stage when information collides during information switching. By this, the cell discard quality can be guaranteed between IWUs. (5) The information extracted from the waiting buffers B33 and B34 is multiplexed by the MUX means B42,
It is sent to the switching means BZ2 via the F means BJ2 and the IWU communication path is selected.

Here, the following algorithm can be considered as a transmission algorithm from the queuing buffer for guaranteeing the cell discard quality. (a) Information can be sent from the queuing buffer that provides QOS with low discard quality only when all the information is sent from the queuing buffer that guarantees QOS with high discard quality. (b) If information is retrieved several times from a queuing buffer that provides a high-quality QOS, a high-quality QOS with a low quality is discarded.
Information can be retrieved from the waiting buffer that provides the.
This number of times can be arbitrarily determined. (c) A queuing buffer that provides a high quality discard QOS is supplied with a faster clock than that sent to a queuing buffer that provides a poor quality QOS, and both buffers are controlled in the same way to the MUX means. Different QOS is provided by sending. (d) A threshold is set in the queuing buffer that provides QOS with high discard quality, and buffer control is performed so that information is not accumulated beyond the threshold.

Here, Q which provides the discard quality and the delay quality
The method of arranging the OS providing means is not limited to the above-mentioned method, and the method of arranging both the QOS providing means of the discard quality and the delay quality in the latter stage of the switching means, or conversely, the QOS providing means both of the switching means. And a method of providing QOS of both discard quality and delay quality by one QOS providing means.

[Embodiment 7] FIG. 33 shows an embodiment of a network connection system using the network connection apparatus according to the present invention. In FIG. 33, from the terminal B661 in the CO & CL mixed LAN (B113) to CO & CL
For the terminal B663 in the mixed LAN (B114),
Connections CO-1, CO-2, CO-3 are set.

Here, the connection CO-1 requested from the terminal B661 is the CO & CL mixed type LAN (B11
3) is set by the connection setting means B441, and CO-3 is set by the connection setting means B442 in the CO & CL mixed type LAN (B114).
Further, the connection CO-2 between the IWU (B223) and the IWU (B224) is set by the connection setting means in the IWU, so Q provided by the CO-2 is provided.
The OS has a configuration that can be arbitrarily set by the connection setting means in the IWU.

In the QOS providing system by the IWU according to the present embodiment, the QOS of the quality which is always higher than the requested QOS in the connection setting request issued from the terminal B661 is provided to the connection CO-2 between the IWUs. It has become. That is, the quality of Q is higher than that of the connections CO-1 and CO-3 that provide the requested QOS.
The OS is provided to the connection CO-2.

By providing the QOS to such a connection, it is possible to reduce the deterioration of the QOS in the IWU communication path. Also, the connection setting means in the network connected by the IWU each independently establishes the QO of the connection in the network.
Even if S management is performed, it is possible to minimize the influence of deterioration of QOS to be provided due to quality deterioration in the IWU communication path.

IW that realizes such a QOS providing method
The QOS providing means in U can be sufficiently provided by the internal means allocating method shown in FIG.

[Embodiment 8] FIG. 34 shows an embodiment of a network connection system using the network connection apparatus according to the present invention. In FIG. 34, terminals B661, B662, B in the CO & CL mixed type LAN (B113) are shown.
663, B664 to CO & CL mixed type LAN (B11
4) Terminals B665, B666, B667, B668 in
, The connections CO-1 to CO-12 are set.

Here, the CO & CL mixed type LAN (B11
3) Connections CO-1 to C requested by the terminals in
O-4 is set by the connection setting means in the CO & CL mixed type LAN (B113), and CO-5 to CO
-8 is set by the connection setting means in the CO & CL mixed type LAN (B114).

Also, IWU (B225) and IWU (B2
26), the connections CO-9 to CO-12 with
Since it is set by the connection setting means in the IWU, the QOS provided by CO-9 to CO-12 can be arbitrarily set by the connection setting means in the IWU.

In the normal QOS providing system, IWU (B2
25) Connection between IWU (B226) and CO
As a QOS providing method in -9 to CO-12, the terminal B is used.
The method of providing the QOS requested from the B.661 to B664 as it is, or the higher quality Q as described above.
A method of providing the OS can be considered. However, in order to provide high quality QOS, it is necessary to set up a connection with a large bandwidth in the communication path between IWUs.
The number of connections that can be set between U may be limited.

When the IWU-provided QOS providing method according to this embodiment is used, it is possible to provide a lower QOS than the requested QOS during the connection between IWUs. By using such a QOS providing method, it is possible to connect between networks by setting a connection that secures a bandwidth that is not too large in the IWU communication path.
It is possible to increase the number of connections that can be set between them, and more terminals can transfer information between IWUs than usual. That is, even a connection for which a request for allocating a band larger than the band (communication capacity determined by the IWU communication path) given between IWUs can be set in the communication path between IWUs.

From this, the communication capacity between IWUs is greater than the communication capacity between IWUs because the number of terminals in the network connected by IWUs increases and the bandwidth of the information sent from each terminal increases. Even if a large amount of communication is required, the quality of service of communication may deteriorate slightly, but it is necessary to cope with changes in communication capacity between IWUs without changing the connection configuration of IWUs or the resources allocated between IWUs. Will be able to.

[Embodiment 9] FIG. 35 shows an example of the format of a packet sent to the IWU when a network connection is made using the network connection apparatus according to the present invention, and the time of the packet sent. An example of changes will be shown. In the IWU according to the present embodiment, the IOS of the QOS according to the characteristics of the packet sent to this IWU is set to the IWU.
It is configured so that it can be set in the U-to-U communication path.

Here, the following characteristics can be considered as the characteristics of the packet, and it can be expected that the following effects can be obtained. (1) Allocating QOS according to the packet length of transmitted information First, when setting a connection between IWUs to transfer information, consider assigning high quality QOS to information with a long packet length. . When transferring information between terminals that are far apart, the probability that a transmission error during information transmission will occur in the packet for information with a long packet length is high, and it is necessary to perform retransmission processing for that purpose. There is a problem that the throughput of a typical packet is reduced. For such a problem, by providing high quality QOS to information with a long packet length, information can be transferred with a low delay time, and thus high throughput can be provided even when retransmission processing is required. become.

Further, by giving a high discard quality to the information having a long packet length, the transmission error rate between IWUs (the transmission error rate including not only the error in the communication path between IWUs but also the information loss due to the cell discard) is increased. It is possible to reduce the number of times the retransmission process is performed.

Since such an effect can be obtained, the method of giving high quality QOS to information having a long packet length is as follows.
In particular, in long-distance information transmission, it is possible to prevent a substantial decrease in information throughput.

As a reverse case of providing high quality QOS for long packet length information, consider assigning high quality QOS for short packet length information. It is considered that the information transmitted from the terminal when providing real-time communication such as a telephone occupies most of the information having a short packet length, so that high quality QOS is given to the information having a short packet length. By providing QOS with this algorithm, it is possible to provide information transfer between networks with low delay and low discard rate for information that requires real-time processing. (2) Information from terminals that have connection setting means, failure avoidance means, or terminals that perform network management as terminals in a communication network that allocates QOS according to control information is used when operating the network. This is very important information. In addition, it is expected that an emergency signal such as a failure detection can be communicated with a transfer delay that allows real-time processing due to its nature.

Since a packet transmitting such particularly important information usually has a bit indicating in the control information field of the packet what kind of information the packet is transferring, It is conceivable to give a high quality QOS to the highly important information from the bit information of the control information. By doing this,
It becomes possible to transfer highly important information within the network with low delay and low discard rate. (3) QOS depending on source address and destination address
As in the case of (2) above, information from a terminal having a connection setting means, a failure avoidance means, or a terminal performing network management is used as a terminal in the communication network by operating the network. This is very important information when going. In addition, it is expected that an emergency signal such as a failure detection can be communicated with a transfer delay that allows real-time processing due to its nature. The above-mentioned method uses the control bit as the bit for judging the importance of the packet. In this case, the QOS is judged by judging whether the source address or the destination address is an important terminal. Will be assigned.

By using such a QOS allocation method, as in the previous case, highly important information can be transmitted with a low delay and
It becomes possible to transfer data in the network with a low discard rate. Further, since the control information is not used, it is not necessary to newly put the control information on the packet having a small header area of the packet such as ATM, and the information in the packet can be effectively used. Further, since it is not necessary to indicate the degree of importance as the control information, it is possible to reduce the means of the terminal for transmitting the information. (4) Allocating QOS according to the burst length of transmitted information As shown in FIG. 35, a case where packets having the same destination information successively arrive is called burst arrival. Here, the burst length refers to the total extension time of the packets that have successively arrived. The QOS allocation method can be realized by replacing the packet length in the method described in (1) above with the burst length as it is.

[0290] Due to such packet characteristics, QOS
In addition to the method of allocating QOS, a method of allocating QOS depending on the distance between IWUs can be considered for the same reason as described in (1) above. For example, a high quality QOS is given to a connection in an IWU communication path connected to a distant IWU, and an IWU connected to a near IWU is connected.
A method of providing low-quality QOS to the connection in the inter-communication path is also conceivable. Further, a method of assigning QOS to each IWU communication path according to the transmission efficiency peculiar to each communication path is also conceivable.

These methods assign QOS to the type of the communication channel or the transmission medium of the communication channel, whereas the conventional methods assign QOS to each connection. Therefore, when IWU is introduced. And connection management means in the IWU when installing or changing the communication path between IWU and IWU
It will be input as information in the OS management means.

In addition, IWU as described above
In contrast to the method of providing high-quality QOS within the IWU for the information and communication paths that are likely to deteriorate in service quality between the IWUs, the information and communication paths that may deteriorate in service quality between the IWUs. Low quality QO
A method of providing S is also conceivable.

By such a method, it becomes possible to connect the network and the information and communication path capable of providing high quality QOS between IWUs without impairing the service quality that can be provided.

Such a QOS providing method is particularly effective when the deterioration of the service quality due to the QOS providing method provided within the IWU as described above is less problematic than the deterioration of the service quality between IWUs. That's the method.

The QOS allocating method as described above is not necessarily applicable only to connectionless type information packets, and is also applicable when transferring connection type information between IWUs. This is a possible method. In addition to the methods described so far, IWU is available as a method for providing QOS for connection-type information.
The distance of the connection itself set between (the distance between the terminals setting the connection) and the number of nodes (hop count) that the connection set between IWU passes through
A method of assigning QOS to

<2> [Embodiment 10] FIG. 36 and FIG. 37 are diagrams showing a schematic configuration of an information communication system according to Embodiment 10 of the present invention.

The information communication system shown in FIG. 36 (a),
Each of the four terminal accommodating packet processing nodes 1001 to 1004 for accommodating a plurality of terminals 1000 and the connection-dedicated packet processing node 1005 for connecting the terminal accommodating packet processing nodes 1001 to 1004 to each other is 155 Mpbs. Is a network connected using four physical links, and the physical network has a multipath at the data link level, but does not have a multipath at the network level. That is, as shown in FIG. 36B, different physical links at the data link level are set to 620 Mbp.
When replaced with one physical link of s, the multipath does not exist.

In the configuration of FIG. 36 (a), the input / output interface speeds of all the physical links are 155M with the same speed.
Although it is set to bps, when a packet processing node capable of reallocating information transfer capacity and accommodating multiple speeds is used as the packet processing node, the terminal accommodating packet processing nodes 1001 to 1004 and the connection dedicated packet processing node 1005 are used. Although there are four connecting physical links between each other, as shown in FIG. 30 (c), a configuration in which an information transfer capacity of 620 Mbps is allocated to one of these may be used for actual information transfer. It is possible.

As described above, when a packet processing node that can reallocate information transfer capacity and can accommodate multiple speeds is used as the packet processing node, it becomes possible to allocate a flexible interface speed to the physical link between the nodes. , The information transfer capacity necessary for the physical link, the information transfer capacity of the physical link itself, and the purpose, the information transfer capacity of the packet processing node can be assigned, and when the network is constructed, system components (nodes, node, It is possible to build a network that can effectively utilize the information transfer capability of (link).

A method of copying and transferring the same information in packet units is also conceivable. By performing such transfer, it is possible to realize information transfer with high fault tolerance against a failure of the connection link.

Further, the plurality of terminals 10 shown in FIG.
37, a network in which four terminal accommodation packet processing nodes (1101 to 1104) for accommodating 00 are connected using four physical links of 155 Mbps, and FIG.
The four terminal accommodating packet processing nodes (11
01 to 1104) are connected to each other using a 620 Mbps 1 physical link, 155 Mbps 2 and 310 Mbps 1 physical link, and a 310 Mbps 2 physical link, respectively, although multipath exists at the data link level. , The network configuration is such that there is no multipath at the network level, and as shown in FIG. 37 (c), different physical links at the data link level are 1 at 620 Mbps.
When replaced with a physical link of a book, there is no multipath, and as shown in FIG. 37 (d), the three networks shown in (a) to (c) can be logically regarded as the same network. Means

Here, when a packet processing node capable of reallocating information transfer capability is used as the packet processing node, the three configurations (a) to (c) shown in FIG. 37 have the same physical connection network. (A network in which terminal accommodation packet processing nodes are connected by four physical links) can be used.

As described above, when there are a plurality of physical links between packet processing nodes, there are a plurality of ways of allocating the information transfer capability for inter-node connection of the packet processing nodes to the physical links. The utilization efficiency of the information transfer capability of the system components changes depending on how the bandwidth of the information transfer capability is allocated to the physical link. Therefore, among the methods for allocating bandwidth to the physical link of the information transfer capacity for connection between nodes, by performing appropriate allocation according to the information transfer capacity required for the physical link, the information transfer capacity of the physical link itself, and the purpose. ,
It is possible to build a network that can effectively utilize the information transfer capability of system components (nodes, links).

Here, as shown in FIGS. 36 and 37, in the case of a network configuration that allows multipath at the data link level but does not allow multipath at the network level as a physical network configuration, packet processing is performed. Since the information transfer path between nodes has multipath only at the data link level, there is no queuing by the buffers of different packet processing nodes, and the queuing in the buffer of the same packet processing node changes the network state. However, the queuing delay time in the buffer inside the packet processing node is
Change all at once.

Therefore, in the information communication system as shown in FIG. 36 and FIG. 37, all the information transfer processing capacities other than the information transfer capacity of the packet processing node necessary for accommodating the terminal among the information transfer capacities of the packet processing node. Is prepared (reserved) as the information transfer capability for connection links between packet processing nodes, and the information transfer capability for connection links between node processing nodes is applied to one (or a plurality of) physical links connected between packet processing nodes at multiple speeds. The maximum bandwidth allocation is allowed while allowing the mixture of packets, and the bandwidth allocation when information is transferred between the packet processing nodes is based on the total remaining bandwidth of one or more physical links existing between the packet processing nodes. And allows information transfer with a transfer band larger than the maximum remaining band of the physical link existing between the packet processing nodes. However, once you measure the difference in delay time between multipath routes, you can judge whether multipath transfer can be used for higher speed transfer or single path transfer can be used for higher speed transfer. You can

For example, considering that the information transfer of a call having the required transfer bandwidth Ws is performed using M paths with the bandwidth of each path being Ws / M, the delay time of the multipath route at the data link level is considered. Is "normalized by the N packet processing time in the internal clock of the packet processing node of the input / output N port" and the minimum transfer delay time is tmin and the maximum transfer delay time is tmax, (tmax-tmin)
When / tmin ≧ M, the single-pass transfer can be performed at a higher speed, and when (tmax-tmin) / tmin <M, the multi-pass transfer can be performed at a higher speed.

As described above, by using the remaining band of the Σinternode connection link, the remaining band of each internode connection link, and the transfer delay time of each path, how to determine the internode transfer band required for a certain call It is possible to judge whether or not high-speed transfer is possible by distributing the information by distributing it to multi-pass or single-pass.

As described above, in the communication system connecting a plurality of packet processing nodes, whether or not to perform the multipath transfer or the single path transfer is determined by using a predetermined conditional expression. Solves the problem of not being able to effectively utilize the information transfer capacity of the network in a path environment (data link level multipath environment), and improves the throughput of the entire system configured by connecting system components (nodes, links) Therefore, it is possible to provide a network that can effectively utilize the information transfer capability of system components (nodes, links).

[Eleventh Embodiment] FIG. 38 is a diagram showing a schematic structure of an information communication system according to an eleventh embodiment of the present invention.

The information communication system shown in FIG. 38 (a),
155M between four terminal accommodation packet processing nodes (1201 to 1204) for accommodating a plurality of terminals 1000
A network in which four physical links of bps are used for connection, and even if a different physical link at the data link level is replaced with one physical link of 620 Mbps as shown in FIG. There are two information transfer paths from the accommodation packet processing node to another terminal accommodation packet processing node, and the physical network has a network configuration in which there are multipaths at the data link level and the network level.

However, the logical network configuration for a certain call of the information communication system is a network configuration in which four terminal accommodating nodes are connected by one logical link of 620 Mbps, and information is used by using the logical network. In the case of considering the transfer, the problem of the information transfer method in the multipath environment at the network level can be replaced with the problem of the information transfer method at the data link level, and the same discussion as in the above-described tenth embodiment. It becomes possible to

Therefore, in a communication system connecting a plurality of packet processing nodes, whether or not to perform multipath transfer or single path transfer is determined by using a predetermined conditional expression. Solved the problem that the information transfer capability of the network below could not be effectively utilized, and system components (nodes, links)
It is possible to provide a network capable of effectively utilizing the information transfer capability of system components (nodes, links) in order to improve the throughput of the entire system configured by connecting the components (see FIG. 38 (c)).

In a multipath environment in which a plurality of packet processing nodes are connected, the physical topology of the network is not defined, and a transfer path for transferring information between End and End exists in the network at the network level. When information is not transferred in the state where multipath is physically or logically eliminated at the network level, the information transfer route is different and queued in different buffers. The maximum transfer delay that can be guaranteed even when the load on the network is extremely high in order to sort information in order at the node that receives information from multiple routes because the variation in delay time between routes is large. The time tqmax and the minimum transfer delay time tqmax when the network load is light are the number of transfer hops of each route and the transfer Calculated from the path length, and based on this information, it is possible to determine how high-speed transfer is possible if the transfer band between nodes required by a certain call is allocated to multi-path or single path and information is transferred. Becomes

For example, considering that the information transfer of a call of the required transfer bandwidth Ws is performed using M paths with the bandwidth of each path being Ws / M, from the multiplex node to the demultiplex node. Minimum transfer delay time tmin
, The minimum transfer delay time from the multiplex node to the demultiplex node when the maximum transfer delay time tmax is "normalized by the N packet processing time in the internal clock of the input / output N port multiplex node" is tqm
in, and the maximum transfer delay time tqmax, (tqmax-
When tqmin) / tqmin ≧ M, the single-pass transfer can achieve higher speed transfer, and (tqmax−tqmin) / tqmin
In the case of <M, high-speed transfer is possible with multi-path transfer.

Therefore, in a communication system connecting a plurality of packet processing nodes, whether or not to perform multipath transfer or single path transfer is determined by using a predetermined conditional expression. Solves the problem of not being able to effectively utilize the information transfer capacity of the network in a network environment in which multipath exists, and improves the throughput of the entire system configured by connecting system components (nodes, links). Therefore, it is possible to provide a network that can effectively utilize the information transfer capability of system components (nodes, links).

As described above, Example 10 and Example 1
According to 1, between the remaining bandwidth of the individual inter-node connection links between the packet processing nodes and the transfer bandwidth required for information transfer between the packet processing nodes, max (remaining bandwidth of the inter-node connection links) <Band required for information transfer between nodes <Σ Even if the relationship between the remaining bandwidth of the connection links between nodes holds, the information transfer capacity of the connection links between nodes does not become a throughput bottleneck, and high-speed packet transfer between nodes It is possible to provide a system in which the ability to transfer information between nodes is less likely to become a bottleneck.

Further, since a plurality of physical links between packet processing nodes are allocated and routed as if they were one physical link, it is possible to simplify the routing route. The cost can be reduced.

Bandwidth allocation when information is transferred between packet processing nodes is assigned to one physical link based on the sum of the remaining bands of one or a plurality of physical links existing between packet processing nodes. As described above, it is characterized by allowing information transfer having a transfer band larger than the maximum remaining band of the physical link existing between the packet processing nodes, so that it has high durability against high-speed and bursty information transfer. A system can be provided.

In addition, there are concentrated multi-paths when information is transferred between packet processing nodes (there is a logical multi-path at the data link level, but there is a logical multi-path at the network level). However, since the number of hops of the multipath route is the same, the problem of delay time variation due to the route and arrival order inversion due to delay variation when using the multipath route at the same time The problem of can be solved easily.

Further, in the information communication system under the multipath environment, it is faster to perform the multipath transfer depending on the internal state of the network (the number of passing nodes, the queuing state, etc.) when the multipath transfer is performed. Solved the problem that it was not possible to determine which transfer method was used to effectively utilize the information transfer capacity, because the case where transfer was possible and the case where single path transfer could be performed at higher speed changed. In addition, in order to improve the throughput of the entire system configured by connecting the system components (nodes, links), it is possible to provide a network that can effectively utilize the information transfer capability of the system components (nodes, links). .

Furthermore, according to the tenth embodiment, since the physical multipath at the network level is not permitted, the physical level at the network level is lower than that at the network level. It has good consistency with the operation and management of the network without allowing the logical multipath, and it becomes possible to utilize the information transfer capability of the system constituent elements (nodes, links) of the network more efficiently.

[0322]

【The invention's effect】

<1> By using the network connection device of the present invention (claims 1 and 2), a flexible network configuration can be taken even when connecting between networks having different types of communication protocols, and the network configuration can be controlled. It is possible to expand or change the network without changing the settings. In particular, by adopting the ATM protocol, which is a connection-type communication protocol, as a unique protocol between network connection devices, it is possible to provide a high-speed and highly reliable communication path between networks.

Furthermore, ATM is installed in the network connection device.
By providing a connection switching means such as a switch, it is easy to identify the network to connect to when connecting multiple networks and to switch the information to be transferred to the network by self-routing with a hardware configuration. It becomes possible to do. Further, since it is possible to allocate a plurality of output interfaces of the switching means in the network connection device to the connected networks by an appropriate number, if the communication bands required between the connected networks are different from each other. Moreover, since it is possible to easily set the communication band between the network connection devices depending on the number of output interfaces to be allocated, it is possible to set the communication band between the networks to the communication band expected by the network designer. . In addition, for the same reason, even if the communication bandwidth required between the connected networks changes with time, the communication path bandwidth between the network connection devices can be easily changed. It is possible to set the communication band of to the state most suitable for the network operation state. Furthermore, by realizing high-speed network connection in this way, when connecting a network for future high-speed broadband communication, it is possible to perform network connection without degrading the performance of the network itself.

Further, according to the present invention (Claim 3), it is possible to set the QOS for each connection even for the connection set between the network connection devices, so that the QOS can be set via a plurality of networks. Even in the connection to be set, it is possible to provide the communication service quality suitable for the connection.

<2> On the other hand, according to the present invention (claim 4), it is suitable for speeding up, routing and route switching control are easy, bottlenecks of connection links are hard to occur, and system components (switch node, connection) It is possible to provide an information communication system capable of effectively utilizing the information transfer capability of (link).

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a network connection system using a network connection device according to a first embodiment of the present invention.

FIG. 2 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 3 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 1-2 of the present invention.

FIG. 4 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 5 is a diagram showing an example of a network connection system using the network connection device according to the example 1-3 of the present invention.

FIG. 6 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 7 is a diagram showing an example of a network connection method using the network connection device according to the first to fourth embodiments of the present invention.

FIG. 8 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 9 is a diagram showing an example of a network connection method using the network connection device according to the embodiment 1-5 of the present invention.

FIG. 10 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 11 is a diagram showing an example of a network connection method using the network connection device according to the embodiment 1-6-1 of the present invention.

FIG. 12 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 1-6-2 of the present invention.

FIG. 13 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 1-6-3 of the present invention.

FIG. 14 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 1-7-1 of the present invention.

FIG. 15 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 1-7-2 of the present invention.

FIG. 16 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 1-8 of the present invention.

FIG. 17 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 2-1 of the present invention.

FIG. 18 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 19 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 2-2 of the present invention.

FIG. 20 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 21 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 2-3 of the present invention.

FIG. 22 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 2-4 of the present invention.

FIG. 23 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 2-5 of the present invention.

FIG. 24 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 2-6 of the present invention.

FIG. 25 is a schematic configuration diagram showing a network connection device according to an embodiment 2-7 of the present invention.

FIG. 26 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 3-1 of the present invention.

FIG. 27 is a diagram showing an example of a network connection system using the network connection device according to the embodiment 3-2 of the present invention.

FIG. 28 is a diagram showing an example of a network connection system using the network connection device according to the fourth embodiment of the present invention.

FIG. 29 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 30 is a schematic configuration diagram showing a network connection device according to a fifth embodiment of the present invention.

FIG. 31 is a diagram showing an example of a network connection system using a network connection device according to a sixth embodiment of the present invention.

FIG. 32 is a schematic configuration diagram showing a network connection device according to the embodiment.

FIG. 33 is a diagram showing an example of a network connection system using a network connection device according to a seventh embodiment of the present invention.

FIG. 34 is a diagram showing an example of a network connection system using a network connection device according to an eighth embodiment of the present invention.

FIG. 35 is a diagram showing an example of the format of a packet sent to the network connection device and the time change of the packet when the network connection device is connected using the network connection device according to the ninth embodiment of the present invention.

FIG. 36 is a diagram showing an example of a schematic configuration of an information communication system according to the tenth embodiment of the present invention.

FIG. 37 is a diagram showing another example of the schematic configuration of the information communication system according to the example;

FIG. 38 is a diagram showing an example of a schematic configuration of an information communication system according to an eleventh embodiment of the present invention.

FIG. 39 is a diagram showing a configuration example of a conventional shared-media LAN system.

FIG. 40 is a diagram showing a configuration example of a conventional star LAN system.

FIG. 41 is a diagram for explaining a throughput neck of a connection link due to the system scale-up in a conventional information communication system.

FIG. 42 is a diagram for explaining a throughput neck elimination by increasing the number of connection links in the conventional information communication system.

FIG. 43 is a diagram for explaining the throughput neck elimination by increasing the interface speed of the connection link in the conventional information communication system.

FIG. 44 is a diagram showing the concept of variable link speed in a conventional information communication system.

FIG. 45 is a diagram showing a multipath transfer system in a conventional information communication system.

FIG. 46 is a block diagram showing the configuration of a conventional information communication system in which multipath exists at the network level.

FIG. 47 is a block diagram showing the concept of network-level multipath transfer in a conventional information communication system.

[Explanation of symbols]

111, 112, A111, A112, A113 ... LA
N, 221, 222, A221, A222, A223 ...
Network connection device (IWU), 31 ... Inter-IWU protocol conversion means, 41 ... Destination information identification means, 51, A
441, A442, A443 ... Protocol conversion means, 7
1, A71 ... Connection management means, 81, A81 ... Destination / connection database, 91, A91-1
~ A91-n ... IWU communication path, A1 ... IWU management CP
U, A881, A882, A883, AM1 ... Connection setting means, A31-1 to A31-n ... Modularized protocol conversion means, AE1 ... Resource management means, AH
1-1 to AH1-n ... Module connecting means, AZ1 ... Switching means, AJ1-3 to AJ1-n ... I / F means, 1000 ... Terminal, 1001-1004, 1101-
1104, 1201 to 1204 ... Terminal accommodating packet processing node, 1005 ... Connection dedicated packet processing node

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical indication H04L 29/06 H04Q 3/00 8732-5K H04L 11/20 B 9371-5K 13/00 305 B

Claims (4)

[Claims] 1. A network connection device provided in each of a plurality of networks for transmitting communication information via an inter-network communication path, wherein each of the network connection devices is the inter-network communication path between the network connection devices. Connection setting / management means for setting an ATM connection to the network and managing the set ATM connection, and for the communication information to be communicated between each of the networks and the inter-network communication path in the network. A network connection device comprising: a protocol conversion means for performing protocol conversion between a predetermined protocol used and an ATM protocol used in the inter-network communication path. 2. A network connection device provided in each of a plurality of networks for transmitting communication information via a plurality of network communication paths, wherein each of the network connection devices is at least one of the network connection devices. Connection setting / management means for setting an ATM connection to the inter-network communication path and managing the set ATM connection, and for the communication information to be communicated between each network and the inter-network communication path. And a protocol conversion means for performing protocol conversion between a predetermined protocol used in the network and an ATM protocol used in the inter-network communication path, and the protocol conversion means and the communication between the plurality of networks. Sui connected to the road Switch means, and should be connected to the protocol conversion means from among the plurality of inter-network communication paths based on the destination information added to the communication information transmitted from each of the networks to the inter-network communication path. A network connection device comprising: a communication path selection unit that selects an inter-network communication path. 3. The communication service quality providing means for providing a predetermined communication service quality to an ATM connection set in the communication path between networks, according to claim 1 or 2. Network connection device. 4. A plurality of packet processing nodes for transmitting and receiving packets to and from at least one of the accommodated terminals, and switching processing of the packets, and a transmission means for interconnecting the packet processing nodes. The packet processing nodes each uniquely determine, for each packet processing node that is a transfer destination of the packet, a combination of the packet processing nodes that relay the packet and an order in which the packet is relayed. Information communication system.