DE4326795A1 - Private communication network with asynchronous transmission mode - Google Patents

Description

The invention relates to a private communication network (or company network) with asynchronous transmission mode (ATM from Asynchronous Transfer Mode) and in particular the ATM network addressing. A solution for that (Layer 2) addressing ports in a private connection-independent (cell relay) virtual network (VPN from Virtual Private Network) with asynchronous over mode with a movable boundary between the Customer Premises Equipment (CPE from Customer Premises Equipment) and the public network and a "seamless" The interface between the two is as follows described subject of the invention. In other words The invention relates to virtual private networks under Ver using an ATM / cell relay (ATM / cell relay) based on permanent virtual circuits in the public network and the indicated movable border. The VPN sees a direct connection-independent Service before, with single cell datagrams in the CPE part and permanent virtual channels (PVCs) in the part of the public network can be used. Architecture from the public network and those supplied by it Services according to relevant international standards are in no way due to the needs of the private Network.

Although the parts of the VPN that are in the public network and implemented in private locations, right different to the different surroundings can use adapted infrastructures Interface between the two actually as "approaching los ". This is largely a result of a very close coordination between the chosen private network address structure and the technology of the case leading to  Establishment of addresses for PVCs on the subscriber network interface places (UNI from User-Network Interfaces) and the network interfaces (NNI of Network-Network Interfaces) in public network is used.

The addressing performance or capability of the ATM Layer of the VPN is through the standard AT cell messages Head size limited to a maximum of 4,096 ports. However, he the header format also allows the addition of 4-way Multiplexers on the access network ports.

According to the invention, cells become free of anything Transported overhead in the information field and it will expects the majority of ports for higher-tier Protocols are designed that address like TCP / IP for a local area network (LAN) coupling or include connection. The number of participants connections that can be operated is as a result very large.

For virtual private networks (VPNs) with ATM / cell Relay use based on the use of permanent of virtual connections / channels (PVCs) in the public Network and a moving boundary between the customer's price equipment (CPE) and the public network with one "seamless" interface between the two begin now to emerge early design specifications.

A PVC is a managed entity and everyone Source port or sender port in the network must be for everyone Destination port to be addressed by it "permanent" access to a PVC. If what any source port on the private network is likely Access to every destination port in the same network is required then the total number of PVCs to be provided the square of the number of sources / destinations what a very large one for a non-trivial private network Represents a significant amount of care in the process Structuring the network requires.

The addressing output is due to the restricted head size, the ATM VPN architectures more than  master everything else, however design specifics tend to tion to assert that "addressing to further study serves ". The described solution optimized the addressing in the private network and then forms this Solution to the PVCs of the public network. It will an architecture and addressing technique described in for whom the illustration is simple and not restrictive.

It is important to note that if the public and customer premise components separately be taken because PVCs are "connection-independent" Provide service, the values in the virtual channel identi identification (VCI of Virtual Channel Identifier) fields and the virtual path identification (VPI of Virtual Path Identifier) fields of the ATM cell header indeed Layer 2 addresses are.

The public network is for all practical purposes open (upgradeable). It has to go through the addressing mechanism no network size restrictions are imposed.

Each location in the private network has a unique, given only address that in the VPI field (UNI format) in upstream direction of an access medium from the public network to the destination address.

Routing is in the intervening public network as normal as for a virtual path using the NNI format. The VCI field is made transparent by the public network.

In the downstream direction on the site The carrier of the destination is the CPI value used the source site address.

According to the present invention, a private Communication network or corporate network provided that an asynchronous transmission mode (ATM) used, having two or more site networks or In short, site networks, which are supported by a public ATM network to be connected, the permanent virtual channels (PVCs) used, the private communication network a connection-independent service between any couple  of connecting ports when addressing them using only the virtual path identification field (VPI) and the vir CCITT's current channel identification field (VCI) recommended ATM cell format and where the full 48 octets of the information field between the Connection ports without one through the ATM layer of the network imposed overhead are transported.

Advantageous developments of the invention are in marked the subclaims.

After that, the VPI and VCI fields lead everyone final ports preferably together the destination address and the source or sender address for a communication cation and the network is divided into subnets, whereby each subnet part of the source and destination address for identification of a connection port within of a subnet and the remaining part of the Source and destination address is used to ensure that Identify subnet.

The local network preferably comprises one or more ATM switching devices or switching devices, each the destination address carried by each cell using a lookup table to interpret the Identity of an outgoing port on the coupling device to provide.

One or more local networks can be a single one Have connection port.

Furthermore, one or more local networks can be one include subnet.

Also two or more of the local networks can each comprise a portion of a subnet (claim 8).

The unused bits of combined VPI and VCI fields can be divided into equally large fields, of which one provides the end point address for a multiplexer, that is connected to the destination port, and that the other the endpoint address for a multiplexer of the Provides source connection ports.

Furthermore, the unused bits can be used to identify one of 2 ⁿ virtual channels (claim 10). In this way, the destination and source identification can also be provided for access to one of 2 ⁿ separate private communication networks.

Preferably a permanent virtual channel in the public network to connect every place with everyone assigned to another location of the private communication network, and the VPI and VCI cell fields leaving a location in the public network access bear the determinants Addresses and source addresses from the private communication network minus the part of the source address that the source subnet identified. Those arriving at one place VPI and VCI cell fields in public network access bear the destination and source addresses of the private communication network reduced by the part of Destination address that identifies the destination subnet certified. By this measure and with a field of at least four bits for identifying the subnet the VPI field on the access bearer of twelve bits eight bits reduced to the generic flow control (FC) field for the subscriber network interface (UNI) lead or carry.

The invention is based on the self understandable drawings explained in more detail. Show it:

Figure 1 is a schematic representation of a hypothetical public / private network.

Figure 2 shows the subscriber-network interface format for an ATM cell addressing.

Fig. 3 is a network-network interface format for an ATM cell addressing;

Fig. 4 is a schematic representation of a typical AMIS network;

Fig. 5 is a schematic representation of a sym metrical network,

Figure 6 shows an address map used within the network of Figure 5;

Figure 7 is a schematic view of a network in which a mixture of small and medium size networks occurs;

Figures 8 and 9 illustrate the address map used within the network of Figure 7;

Fig. 10 shows the address map used with a single network port;

Fig. 11 shows a variant of the address mapping of Fig. 6, indicating the difficulties that arise when there is more than one mini network; and

Fig. 12 shows a modification of the network for overcoming the problem indicated in Fig. 11.

Figs. 1 shows two separate connections of a public network in a hypothetical network: Coventry to Beeston and Beeston to Poole.

The received VPI and VCI values are in the Customer premise equipment used the source location or source location in the private network as well as the determ address at the destination or destination along with the source location source address certify. A more complete description of the address Mechanism of adjustment follows later.

Unlike a public network, a private one Network have a defined size limit, and it can take advantage of reduced costs and complexity be exploited. A concept that takes advantage of this Fact has been developed. The concept will as AMiS (from ATM mini-switches = ATM mini-coupling or Switches) and is the counterpart was the patent application No. GB 91 08 824.5, the open content for the definition and definition of this con included in the present application is.

Absolute addresses for all ports of the assigned private network and the identification and Source addresses are in the VPI and VCI fields because every ATM cell using the NNI format guided. The maximum number of network ports served is based on the 12-bit VPI field size Limited to 4,096. However, this number can be wagered through  expanded by 4-way multiplexers on suitable network ports become. There is no over in the information field of cells head led, which is why adjustment or adaptation layer information (e.g. the TCP / IP datagrams for a LAN connection or coupling supports it accepted format apart from a segmentation and Reunification between packets and cells without modifi cation can use.

The 4-bit field used by the multiplexer as above mentioned is divided into two 2-bit fields, one of which provides the destination port address of the multiplexer and the other is the sender or source port address of one remote sender or source multiplexer (counter multiplexers).

The UNI and NNI formats for ATM cells are shown in Figures 2 and 3 along with the use of the fields for AMiS addressing as described here.

A schematic representation, which shows an example of a typical AMiS network, is shown in FIG . Two uses of the remaining four bits of the VCI field are shown, namely to provide source and destination addresses for a 4-way multiplexer and to provide up to 16 predefined or negotiated virtual channels (VCs) between a pair of network ports. If one end of the connection is a multiplexer, the number of VCs served is reduced to four.

For a network that would cover some locations part of the above model including part of the Coupling or switching from the public network provided. Using the procedure of the part described in the previous part The addressing of the public network becomes cells transported transparently. The private network is out of order the interface between the public network and the Site CPE from using public connections unaffected network.

Since the destination address in the VPI field is a is "absolute" address, it is in individual coupling  or switching facilities not translated, but a fold on the outgoing port on the switching equipment interpreted and forwarded unchanged.

Let us now consider a symmetrical ("balanced") network in which AMiS is mapped onto the public network, this being the simplest case for a network in which each location comprises an AMiS mini network. The public network provides the cross-switching between the local networks, as shown in Fig. 5.

Within the local network, the full message head format is used, as shown on the left in Fig. 6. On the access trunk of the public network, the four bits used to identify the destination mini network are replaced by the GFC field (flow control field), whereas both the source and destination mini network identities are implicit in the PVC identity.

The illustrations at points A and B in FIG. 5 are also shown in FIG. 6.

The destination mini network address is used for this det, the PVC of the public network at the UNI des Quel len mini-network, and the source mini-network Address is used for the PVC on the destination mini network to identify. The format of the remaining fields is designed to be a general solution that is of size of the local area network is unaffected, to enable how further is described below.

In the core network are the 16 PVCs for each location are required using 16 out of 256 possible of the NNI format. The rest of the header address fields is transported unchanged by the network.

One of the 16 destination addresses becomes the address the source or sender location, and it is not to expect a location to address itself, so this address can be redundant. Research the network changes at least one reserved CPI value for his own needs, however, it will ensure that the location  address coincides with the reserved address, then 16 locations can still be served. Since the reserved VPI value is fixed and the location address is variable, a small CPI value corrector or -Translator required at the site interface.

In Fig. 7, a medium size location is Darge includes a mini network (B), which is connected by the public network with a number of small locations, each comprising a segment of a mini network.

Two small locations (A1 and A2) are shown on the left side of FIG. 7.

In this case, the address mapping is very similar to that of the previous example. For A1 or A2 to B, the illustration corresponds to the diagram shown in FIG. 8.

Note that the same cell header formats as before, using only the limit at which the public network has an interest is changed.

Using the same universal cell formats can create a single network port by moving the boundary of the public network can be handled by another four bits. This implies the expansion beyond the border of the CPI Into the most significant (MS) 4 bits of the field VCI field.

A site that spans more than one mini-network cannot be served as a single carrier because the available address space is too limited. This fact is shown in Fig. 9 for a variant of the first example.

The solution to this problem consists in separate cross connections from the local and public network, as shown in Fig. 10.

The cross-connection from the public network shows the downstream side subscriber network interfaces on. Since the UNIs have generic flow control (GFC), they must be in the handling system from the public network be included. Therefore, the public network cross belongs connection probably the operator of the public Network, although it is on the customer side.

Claims (11)

1. Private communication network that is an asynchronous Transmission mode (ATM) used, having two or local networks more connected by a public ATM network using permanent virtual channels (PVCs), the private communication network being a connection independent service between any pair of addressers port ports using only the virtual path identification field (VPI) and the virtual channel id certification field (VCI) of the CCITT recommended ATM Cell format provides and the full 48 octets the information field between the connecting ports without an over imposed by the ATM layer of the network be transported head. 2. Network according to claim 1, characterized, that the VPI and VCI fields are common to all connection ports sam the destination address and the source address for one Bear communication and that the network is broken up into subnetworks is shared, with each subnet part of the source and destination address to identify a port ports used within a subnet and the lead part of the source and destination address is used to identify the subnet. 3. Network according to claim 2, characterized, that the local network one or more ATM switching devices comprises and that each switching device input the determ address that is kept by each cell by means of a lookup table, interpreting the identity an outgoing port provided on the coupling device becomes.   4. Network according to claim 2, characterized, that one or more local networks have a single connection port exhibit. 5. Network according to claim 3 or 4, characterized, that the network comprises a single local network. 6. Network according to claim 3 or 4, characterized, that it has two or more local networks that are perma nent virtual channels (PVCs) are connected to each other, which are managed by the public network, and that a permanent virtual channel in the public network for ver binding every place with every other place of private Communication network is assigned and that the one place leaving VPI and VCI fields of cells in public Network access bearers the destination and source addresses from the private communication network reduced by the part of the Maintain source address that identifies the source subnet and that the VPI and VCI fields from Cells in the public network access carrier the destination and source addresses of the private communication network ver reduced by the part of the destination address that the Destination subnet identified, through this Measure and with a field of at least 4 bits for Identification of the subnet the VPI field on the access carrier from 12 bits to 8 bits is reduced to the generi flow control (GFC) field for the subscriber network interface agency (UNI). 7. Network according to claim 6, characterized, that one or more of the local networks is a single subnet include.   8. Network according to claim 6, characterized, that two or more of the local networks each have a share of one Subnet and that two or more of these share Subnets through permanent virtual channels (PVCs) of the public network for training a single part network are connected. 9. Network according to claim 7 or 8, characterized, that any remaining and unused bits of the combined The VPI and VCI fields were of the same size in two fields are divided, one of which is the end point address for a multiplexer that provides with the determ port of the network is connected, and the other provides the endpoint address for a multiplexer that uses the source port is connected. 10. Network according to claim 7 or 8, characterized in that any remaining and unused bits of the combined VPI and VCI fields are used to identify one of 2 ⁿ virtual channels between two connecting ports, where "n" is the number of remaining Bits is. 11. Network according to claim 7 or 8, characterized in that any remaining unused bits of the combined VPI and VCI fields are divided into two equally sized fields, the identification and source identifications for the public network to access one of 2 ⁿ separate private Provide communication networks.