WO2002035860A1 - Adaptive regulation in a mobile system - Google Patents

Abstract

The present invention relates to an arrangement and a method for Automatic Congestion Control. According to one embodiment of the invention the traffic type that causes congestion between switching nodes in a mobile telephone system is registered. Further information about the traffic type is sent from the overloaded switching node to the overloading switching nodes and a reduction of that traffic type takes place according to pre-defined methods.

Description

ADAPTIVE REGULATION IN A MOBILE SYSTEM

TECHNICAL FIELD OF THE INVENTION

The present invention relates generally to a mobile telephony system and, in particular, to a system having automatic congestion control as well as a method for the same.

DESCRIPTION OF RELATED ART

The Personal Digital Cellular (PDC) system is a mobile cellular radiotelephone system that has been introduced as a standard in Japan. The basic telecommunications services supported by the PDC standard include two-way voice communications between fixed and mobile subscribers, a Short Message Service (SMS) and data transmissions.

Networks designed to handle normal traffic load can become overloaded and congested when there is an unusual increase in traffic. Network management reduces the negative effects of overload and faults in the network through efficient utilisation of network resources and capacity. A switching subsystem or functionality that supervises the traffic flow through the exchange and introduces temporary changes in that flow, such as a Network Management Subsystem (NMS ) , includes some means for overload control. One such control system is the call-gapping algorithm disclosed in U.S. Patent number 5,060,258 entitled ^vCall Traffic Control'.

SUMMARY OF THE INVENTION

The use of various methods and algorithms for overload control alleviates the problem of congestion in peak time traffic. However, the methods used for overload control this far are undiscriminating and indifferent to the type of traffic that is causing the congestion; only information about the overload level is transmitted between adjacent switching nodes. Thus, the congestion control methods this far may severely limit the traffic of one kind although the congestion is caused by traffic of another kind. For example, in Japan the PDC system has SMS capabilities that due to the inexpensive use is much used by its subscribers; very often the SMS traffic corresponds to around 50% of the overall traffic in the networks. Because of this the SMS traffic frequently causes congestion in the networks.

The present invention alleviates the present limitations and shortcomings of congestion control by pointing out the cause of the congestion and further posing limitations discriminatingly .

In accordance with a first aspect of the invention an object is therefor to provide means and a method for registering the various types of traffic that is coming to one switching node from another switching node.

Another object of this invention is to use the information about traffic types to further control the traffic at traffic congestion and efficiently alleviate such conditions.

DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates a schematic configuration showing various external networks connected through a couple of TMSC to a VMSC (prior art) .

Figure 2 illustrates a more detailed view of two connected different level MSC - one TMSC and one VMSC - including LOAS in each MSC (prior art) .

Figure 3 illustrates an exemplary register for ISUP traffic in the LOAS.

Figure 4 illustrates a flow chart for the ACC in a preferred embodiment . DESCRIPTION OF PREFERRED EMBODIMENTS

The present invention aims at alleviating problems with traffic congestion when various types of traffic are present in and between the various networks .

For communication between nodes within a communication network and for communication between different networks, various protocols have been designed to meet the special requirements that are posed on such flow of information. Two such protocols that have been standardised and implemented in various systems are the ISDN User Part (ISUP) and the Mobile Application Part (MAP) .

ISUP is a protocol that allows signalling for routing, setup and release of calls between exchanges in the digital phone network, such as PDC, and exchanges in the fixed network. In Japan a number of different ISUP protocols are used to interconnect different networks.

MAP is a protocol specially designed to support the call control and mobility functions in a cellular system. This protocol is implemented in, for example, the VMSC, TMSC and HLR so that these nodes can communicate with each other when processing location registration, location cancellation, handover, transfer of authentication keys and other functions that require interaction with several systems simultaneously.

Now referring to figure 1 that shows a schematic of a Public Land Mobile Network (PLMN) including exemplary switching nodes within the PLMN as well as connected external networks of various kinds .

A Transit Mobile Switching Centre (TMSC) 10 is a Mobile Switching Centre (MSC) without connecting Base Stations (BS) 500. Thus it acts as a transit switch routing traffic towards one or several external networks, such as Public Switched Telephone Network (PSTN) 100, Public Land Mobile Network (PLMN) 300, and Short Messages Cervices Centre (SMS-C) 400, and the Visiting Mobile Switching Centre (VMSC) 20 that it services. Incoming voice calls from PSTN 100, PLMN 300 and short messages from SMS-C 400 to the TMSC 10 result in an interrogation to the Home Location Register (HLR) 70 about the location of the Mobile Station (MS) 600 and then in a transit routing of the voice call or short message to the corresponding VMSC 20. The voice call or short message is further transferred to one of the BS 500 in communication with the MS 600.

Further, every TMSC 10 is normally connected to a number of PLMN 300 and SMS-C 400, and several different PSTN 100 networks. For redundancy reasons two TMSC 10 are normally connected to one and the same PSTN 100 SMS-C 400 or PLMN 300; if one of the TMSC is not accessible at a certain time from one of these 'external' networks (PSTN, SMS-C or PLMN) the other TMSC can be used instead.

Also, one and the same TMSC 10 is normally connected to all VMSC 20 in the same operator network (same PLMN) . For example, if the operator of the network has ten VMSC 20, then every TMSC 10 in the same operator's network is connected to one and each of these ten VMSC 20. Each VMSC 20 is further connected to a number of BS 500 with which the mobile terminals or mobile stations (MS) 600 communicate. A call from mobile station A in VMSC1 to mobile station B, operating in another area covered by VMSC2 , is connected through one of the two TMSC, setting up a call MS (A) - BSla - VMSC1 - TMSC1 or TMSC2 - VMSC2 - BS2b - MS(B). Thus, the network is structured in several layers.

Operators having small or limited networks including less than four MSC in the network usually have all the MSC of the network directly connected to each other in a meshed configuration. Larger networks would require too much transmission in such a configuration, why another layer is introduced and transmission limited to occur between different layers. However, some VMSC ' s with a lot of traffic between them may have a direct connection between them also in a larger network in order to save TMSC Central Processor capacity.

There are normally several HLR nodes 70 in the same network, due to the number of subscribers that the HLR 70 can maximally register in its database. Frequently more subscribers are registered in one and the same network than can fit into one HLR 70. If that is the case all HLR 70 are connected to all TMSC 10 and VMSC 20 in the same network.

In a preferred embodiment all signalling from a mobile station (MS) 600 is received by at least one BS 500, transmitted from the BS 500 to a connected VMSC 20 and further to one of the two connected TMSC 10. The mobile terminal originating call accesses the HLR 70 from the originating VMSC 20.

If the call is made to a mobile terminal within the same PLMN the VMSC 20 receiving the call from one of the connected BS 500 accesses a connected HLR 70 to request the location of the B-terminal (call receiving terminal) . The HLR 70 responds with the B-subscribers VMSC 20. If the B-terminal resides within the same VMSC 20 then the call is transmitted via the appropriate BS 500. But if the B-terminal is located within another VMSC the call is transmitted via one of the TMSC 10 to the appropriate VMSC 20.

If the B-subscriber is in one of the 'external' networks, for example another PLMN 300 or a PSTN 100, then the HLR 70 is not accessed but the call is subsequently routed from the TMSC 10 receiving it from the VMSC 20 directly to the B- subscriber's PLMN 300 or PSTN 100. An address tag from the originating TMSC 10 indicates the gate to the terminating B- subscriber's network. In the case the call consists of a Short Message the TMSC 10 transmits the message to the SMS-C 400.

Normally every SMS-C 400 is connected to two TMSC 10 for the same redundancy reasons as in the case of the PSTN 100 and PLMN 300 networks. The same carrier and signalling protocol (ISUP) is used in the SMS case, the only difference being that the SMS message is included in the ISUP message itself. Apart from that the connection scheme is the same as in the case of a voice call.

If an incoming call to a MS 600 in the network is received at one of the two TMSC 10, a request is sent from the TMSC 10 to the HLR 70 about the location of the MS 600. The HLR 70 responds with the VMSC 20 identifier and the TMSC 10 routes the call to the appropriate VMSC 20 for further transmission via one of the BS 500 to the MS 600.

Now referring to figure 2, there is a TMSC 10 connected to external networks consisting of PSTN 100, PLMN 300 and SMS-C 400. Further the TMSC 10 is connected to a VMSC 20, and both the TMSC 10 and the VMSC 20 are connected to a HLR 70. The VMSC 20 is also connected to a number of BS 500. Traffic is bi-directional between the two switching nodes 10 and 20. Like numbers in figures 1 through 3 are used to represent the same software or physical units.

As explained earlier, although figure 2 only shows one VMSC 20 and one TMSC 10, it is commonly the case that each VMSC 20 is connected to two TMSC 10 and each of the two TMSC 10 is connected to all VMSC 20 of the network. Also, there may be several external networks 100, 300 and 400, all of them connected to both TMSC 10 of the PLMN in question. Furthermore, there are normally several HLR 70 in the network, all of which are connected to both the TMSC 10 and all the VMSC 20 of the same network. Also, each switching node TMSC 10 and VMSC 20 has a function block called Load Regulation Services (LOAS) 30 and 40. The LOAS 30 and 40 control the traffic processing in relation to available Central Processor (CP) capacity. If needed LOAS 30 or 40 will buffer, due to CP load limit, requests for locally originated calls, incoming calls, subscriber services that generate new calls, trunk calls like connection of answering machines, and miscellaneous calls, which activities are not directly related to call setup, until the CP capacity is available again. By this LOAS 30 controls the traffic intensity in relation to the capacity available.

Now reference is made to figure 3. In an exemplary embodiment each TMSC 10 and VMSC 20 in the network contains a function block LOAS 30 and 40. LOAS keeps track of the load situation in the switch. Further, in a preferred embodiment, LOAS will also register load according to traffic type. With traffic type is meant capacity priority in VMSC 20 or TMSC 10. LOAS 30 (or 40) processes traffic requests according to the priority that is assigned to it, as further explained later. If one of the switching nodes VMSC 20 or TMSC 10 experiences overload LOAS sends an overload indication to adjacent nodes according to the PDC standard described in TTC and RCR 27. In the following example the VMSC 20 is taken as the overloaded node and the TMSC 10 as the overloading node. It is easily understood that the situation may be the reversed. In the case of small networks or otherwise directly connected MSC (VMSC) , this is applicable to traffic between adjacent nodes of the same kind and in the same layer.

The overload with traffic type indication is initiated from the LOAS 30, via Call Control (CC) 34 and sent to ISUP 36 where it is included in the next ISUP release message of the overloaded switching node, here the VMSC 20. The overload with traffic type indication is extracted from the ISUP release message by ISUP 46 and sent to the Network Management System (NMS) 48 of the overloading switching node or nodes, for example one of the two TMSC 10 in the exemplaχ-y embodiment. Two states of overload are defined - overload and severe overload. The adjacent nodes respond to this release message overload indication by reducing traffic towards the overloaded node using the pre-defined reduction method most suitable to the overload situation. The method is pre-defined in a response program in the NMS .

There exist various response programs for traffic reduction such as temporarily buffering or rejecting traffic as well as rerouting traffic via other switching nodes. Rerouting can be applicable when one of the TMSC nodes 10 experiences overload. If that is the case an overload message is sent in an ISUP 'release' message from the overloaded TMSC 10 to the VMSC 20 and the VMSC 20 may reroute outgoing traffic to the other TMSC 10, thus reducing the traffic to the overloaded TMSC 10.

Figure 3 further shows an exemplary configuration of the LOAS 30 having m capacity request buffers 31, each buffer holding up to n capacity requests 33. Each buffer represents a certain priority level. For example, if m=15, then the 16 buffers (0, 1, ... 15) represent different priority levels, where for example buffer 0 is assigned the highest priority. LOAS 30 and 40 process buffers in order of priority; therefor traffic with low priority will first reach a buffert level for Automatic Congestion Control (ACC) in the release message. Information about processor load is sent from the Central Processor (CP) 32 or 42 in the switch to the corresponding LOAS, thus from CP 32 to LOAS 30 and from CP 42 to LOAS 40. Part of the buffers are used for normal incoming calls (voice and SMS) and may be defined, according to this invention, so that every external network PSTN 100, PLMN 300 and SMS-C 400 is represented by one or several corresponding buffers, each buffer given its specific priority. Then, again according to this invention, a call request from, fox- example, the external network PSTNl is registered in, for example, buffer" number" 1, and a call request from PLMNx is registered in buffer number q. Another call request from the same external network may have another priority and is subsequently placed in another buffer, say for example r. Calls originating within the same network are registered in other buffers in the same LOAS.

According to one embodiment of this invention not only the different external networks PSTN 100, PLMN 300 and SMS-C 400 may be given different priorities and calls from within the same such network assigned different priorities depending on the kind of call, for example different priority for voice and data; also the different VMSC's 20 may be assigned different priorities in the LOAS 40 of the TMSC 10. In the same manner the two different TMSC's 10 within the same PLMN may have different priorities in the LOAS 30 of the VMSC 20. The same applies to VMSC's 20 that are directly connected to other VMSC's 20 in small or some bigger networks.

Now referring to figure 4 that shows a preferred embodiment according to this invention exemplifying the case when a traffic type is causing overload. Reference should at the same time be made to figure 3 for the locations of the various steps, observing that the numbering in figure 4 is not co-referential with the numbering in figure 3.

In the first step 11 of this algorithm a Request for capacity is received from the Call Control subsystem (CC 34 or 44 in figure 3) to the LOAS (30 or 40 in figure 3). With the request for capacity is included information about the type of traffic (corresponding to source - internal or external PLMN, PSTN, etc) that it represents.

In the second step 21 the LOAS (30 or 40) checks the Central Processor (CP) of the switching node (VMSC 20 or TMSC 10) . The third step 31 is conditioned, so that if and when the request for traffic type is causing overload (CP load>loadability limit), then step 51 follows. Otherwise, if CP load is less than the loadability limit, the LOAS (30 or 40) responds to CC (34 or 44) with 'capacity granted' .

In step 51, when the loadability limit is passed, the LOAS 30 or 40 registers the request in the capacity request buffer indicated by traffic type.

In step 61 IF the buffer usage is greater than the buffer loadability limit, then the procedure passes on to step 71. Otherwise it is returned to step 11 described previously.

In step 71 the overload response from LOAS (30 or 40) to CC (34 or 44) includes information about the traffic type that causes the overload.

In step 81 of this exemplifying algorithm CC (34 or 44) informs the ISUP handling software (36 or 46) that overload is present and which traffic type that caused the overload.

In step 91 ISUP places an Automatic Congestion Control (ACC) overload indication in the next release message sent to the overloading MSC node (VMSC 20 or TMSC 10) indicating the traffic type causing the overload.

Finally, in step 101 the overloading MSC node (VMSC 20 or TMSC 10) responds to the release message from the overloaded MSC node (another of the VMSC 20 or TMSC 10) by a predefined response program in the NMS (38 or 48) , resulting in a reduction of the overloading traffic type.

Although the invention is described in the present way it is easily seen by anyone skilled in the art that numerous rearrangements and changes may be done without departing from the gist of this invention.

Claims

1. Mobile communication system having more than one switching node where at least one switching node experiences overload, c h a r a c t e r i s e d in that an overload message including information about traffic type causing said overload is sent from said overloaded switching node to the overloading adjacent switching node .

2. Mobile communication system according to claim 1, c h a r a c t e r i s e d in that said information about traffic type causing overload results in reduction of said traffic type from said overloading adjacent switching node to said overloaded switching node.

3. Mobile communication system according to any of claims 1 or 2, c h a r a c t e r i s e d in that said traffic type is ISUP or MAP.

4. System for countering overload conditions in a mobile communication system having more than one switching node, c h a r a c t e r i s e d in that an overload message including information about traffic type causing overload is sent from said overloaded switching node to the overloading adjacent switching node.

5. System according to claim 4, c h a r a c t e r i s e d in that said information about traffic type causing overload results in reduction of said traffic type from said overloading adjacent switching node to said overloaded switching node.

6. System according to any of claims 4 or 5, c h a r a c t e r i s e d in that said traffic type is ISUP or MAP.

7. Method in a mobile communication system having more than one switching node where at least one switching node experiences overload, c h a r a c t e r i s e d in sending infoxmation about traffic type from the overloaded switching node to the overloading switching node.

8. In a system for countering overload conditions in a mobile communication system having more than one switching node, a method for responding to overload c h a r a c t e r i s e d in sending information about traffic type from the overloaded switching node to the overloading switching node.

9. The method of claim 7 or 8, c h a r a c t e r i s e d in reducing the traffic type causing overload from the overloading switching node to the overloaded switching node .

10. The method of any of claims 7-9, wherein said traffic type is ISUP or MAP.