WO2005032089A1

WO2005032089A1 - A distributed processing system and method using h.248 protocol on mgc/mgw

Info

Publication number: WO2005032089A1
Application number: PCT/CN2004/000764
Authority: WO
Inventors: Zichun Zhou; Chaofeng Zhang; Guang Li; Yuan Lu
Original assignee: Zte Corporation
Priority date: 2003-09-28
Filing date: 2004-07-07
Publication date: 2005-04-07
Also published as: CN1492697A; CN1210926C

Abstract

The invention discloses a distributed processing system and method using H.248 protocol on MGC/MGW. The system includes one or more H.248 interface boards, which receive external H.248 messages, identify and distribute the H .248 messages to their respective processors; one or more H.248 distributed processors which process the messages from the H. 248 interface boards, generate the relative H. 248 messages and then transmit them; H. 248 centralized maintenance processor, which processes the error messages from the H. 248 interface boards or the maintenance messages for the MGC/MGW and maintains the normal communication between MGC and MGW This invention resolves the problem that H. 248 protocol can't enlarge the capacity smoothly.

Description

H. 248 protocol distributed processing system and method on MGC / MGW

Technical field

The present invention relates to the field of communications, and in particular, to the large-capacity H.248 protocol distributed processing system and method. Background technique

The H.248 protocol is adopted by organizations such as 3GPP / 3GPP2 as a separate protocol standard for NGN network bearer and control. It is mainly used for gateway control and bearer establishment control between MGC and MGW.

The MGC communicates with the MGW through H.248 messages. These messages can be encoded in text or binary mode. H.248 message length is usually between 100 and 1000 bytes. The format of the unencoded H.248 message is shown in FIG. 1. From the outside to the inside, the format is Transaction, Action, and Command, where: Transaction: A transaction consists of several actions. A transaction has For completeness, actions and commands in the same transaction require sequential execution.

Actions: Actions are composed of several commands. Actions generally have specific meanings. In special cases, an action consists of a Chennai command. An action is actually an identification of a set of commands.

Commands: Commands are the basic unit of interactive information between MGC and MGW. Commands include various command parameters. The operation object of the command is the context (Context), and the context is distinguished from the context identifier (Context ID). The context corresponds to the call and maintains the status of the terminations participating in the call.

A complete call process will interact between MGC and MGW 20 to 60 times H.248 Command to complete the call control operation on the same context, where each interactive command is encapsulated into a transaction to form an H.248 message and sent to the other party.

The currently commonly used H.248 protocol processing scheme is shown in Figure 2: The MGC or MGW has one or more H.248 interface boards externally, which is responsible for receiving external H.248 messages and sending H.248 messages to the outside; received The H.248 message is sent to the H.248 protocol processing board for decoding, and the corresponding H.248 transaction processing, context maintenance, and encoding of outgoing messages are performed on the board. In this solution, the entire system has only one processor for protocol processing, so it is greatly restricted by the processor's capabilities. Even if the processor has a strong capability, it is difficult to cope with the large-capacity H.248 protocol. Taking the PENTIUM III700M CPU as an example, the processing capacity of the H.248 protocol of the above scheme is usually less than 100 calls / second.

In the NGN system, there is a demand for MGC processing capacity of more than 1000 calls per second (such as 3GPP's MSC-SERVER), and the simple single board processing of the H.248 protocol can no longer meet the system requirements. The above simple processing method cannot be used to expand the processing capacity of the system by adding processors, because it is not known which processor the message corresponds to before the H.248 message is decoded. If the interface board arbitrarily sends the received H.248 messages to each processor for decoding, this will inevitably lead to a large number of decoded messages that do not belong to the processor. Assuming there are η (η> 1) processors in the system, from a probability perspective, among them (l-1 / n) * 100% of messages need to be forwarded, and each message is forwarded between 1 and n- Between 1. The more processors, the greater the probability of message forwarding, and the more times each message is forwarded to the final context processor. Based on a model with 10 processors processing 1000 calls per second, 90% of messages are forwarded. Frequent message forwarding will consume a lot of processor resources and internal communication bandwidth. In addition, since the length of the H.248 message is not fixed, the length of the decoded message is also uncertain, which brings great difficulties to the allocation and management of the communication data area. At present, no solution to the above problems has been proposed. Summary of the invention

An object of the present invention is to provide a distributed processing system and method for the H.248 protocol that can smoothly expand the processing capacity of the system, so as to process the large-capacity H.248 protocol in order to solve the NGN system. The problem of load sharing in H.248 protocol. Among them, the expansion of the system capacity is smoothly achieved by expanding the number of H.248 distributed processors and H.248 interface boards.

In order to achieve the above objective, according to one aspect of the present invention, a distributed processing system of H.248 protocol on MGC / MGW is provided, including: one or more H.248 interface boards, each of said H.248. The 248 interface board is configured to receive external H.248 messages, identify and distribute each H.248 message to its corresponding processor; one or more H.248 distributed processors, and each of the H.248 distributed processing The processor is configured to process the H.248 message distributed by the H.248 interface board, and generate a call-related H.248 message to be sent out; the H.248 centralized maintenance processor is configured to process each H.248. The error message sent by the 248 interface board and the maintenance message for the MGC / MGW are processed, and the normal communication between the MGC and the MGW is maintained.

Preferably, each of the H.248 interface boards further pre-decodes the received H.248 message; the pre-decoding is incomplete decoding.

Preferably, each of the H.248 distributed processors has a MGC / MGW system resource interval allocated to it.

Preferably, when the system is on the MGC, a system resource interval of each of the H.248 distributed processors includes a transaction ID interval and an event request ID interval allocated to the MGC; when the system is on the MGW The system resource interval of each H.248 distributed processor includes a transaction ID interval and a context ID interval allocated to the MGW.

Preferably, each of the H.248 interface boards has a comparison table between the MGC / MGW system resource interval and each H.248 distributed processor; each of the H.248 interface boards identifies the corresponding table according to the comparison table. The H.248 distributed processor corresponding to the received H.248 message, and distributes the H.248 message to the H.248 distributed processor for processing.

According to another aspect of the present invention, an H.248 protocol distributed processing method using the H.248 protocol distributed processing system on the MGC / MGW is provided, including: each of the H.248 interface boards receives An external H.248 message, identifying the H.248 message and distributing it to a corresponding processor; each of the H.248 distributed processors receiving and processing the H distributed by each of the H.248 interface boards .248 messages, and generate call-related H.248 messages to send outward; the H.248 centralized maintenance processor receives and processes error messages sent by each of the H.248 interface boards and Maintenance messages for this MGC / MGW.

Preferably, each of the H.248 interface boards has a comparison table between the MGC / MGW system resource interval and each H.248 distributed processor, and the identifying and distributing step further includes: the pair of H.248 interface boards The received H.248 message is decoded; the H.248 interface board determines the processor corresponding to the H.248 message according to the decoding result and the comparison table; the H.248 interface board converts the H.248 message. A 248 message is sent to the determined processor.

Preferably, the determining step further includes: judging the decoded H.248 message, and if the H.248 message is a transaction response message sent by an external entity to the MGC / MGW, then the decoded transaction is used. The ID looks up the comparison table to determine the H.248 distributed processor to which the H.248 message belongs; if the H.248 message is a transaction request from an external entity to the MGC / MGW, then when the system is on the MGC When searching the comparison table according to the decoded event request ID to determine the H.248 distributed processor corresponding to the H.248 message; when the system is on the MGW, searching the comparison table according to the decoded context ID To determine the corresponding H.248 distributed processor in the range where the context ID is located, if the context ID is new or empty, specify an H.248 distributed processor, and the H.248 distributed processor is the context A value is assigned to the ID; if the H.248 message is decoded to obtain an error message and a maintenance message for the MGC / MGW, the H.248 message is sent to the H.248 centralized maintenance processor.

Preferably, the method further includes: after each H.248 distributed processor finishes processing the content belonging to the processor in the H.248 message, determining whether there is any content that needs to be processed by another processor, and if so, then Distribute the H.248 message to the relevant processor and wait for the processing result; after processing all the content in the H.248 message, the H.248 distributed processor judges whether the H.248 message is other processing If it is forwarded by the router, if it is, it will reply the processing result to the source processor; if not, it will encode the processing result into an H.248 message and send it.

Preferably, the method further comprises: the H.248 centralized maintenance processor further receives and processes a maintenance request message from each of the H.248 distributed processors, and notifies the H.248 distribution of the completion of the processing. Processor.

By adopting the H.248 protocol distributed processing system and method of the present invention, after calculation, Compared with the existing technology, it can solve the problem that the H.248 protocol cannot smoothly expand the capacity. The H.248 call processing capability it can support is at least 20 times higher than the current one. It is the key technology to solve the large-capacity GW and MGC in NGN systems. breakthrough. BRIEF DESCRIPTION OF THE DRAWINGS

The specific implementation of the present invention will be further described in detail below with reference to the accompanying drawings.

Figure 1 is a schematic diagram of the structure of an H.248 message;

FIG. 2 is a structural diagram of a simple processing mode of a conventional single-processor H.248 protocol;

FIG. 3 is a schematic structural diagram of an H.248 protocol distributed processing system on the MGC / MGW of the present invention;

FIG. 4 is a flowchart of the interaction between the main processing links of the H.248 message in the H.248 protocol distributed processing system on the MGOMGW of the present invention;

FIG. 5 is a schematic diagram of the processing flow of the H.248 message pre-decoding and distribution link in FIG. 4; FIG. 6 is a schematic diagram of the processing flow of the H.248 message distributed processing link in FIG. 4. detailed description

Figures 1 and 2 have been described in the background art, and are not repeated here.

FIG. 3 is a structural diagram of the H.248 protocol distributed processing system on the MGC / MGW of the present invention, including: N H.248 interface boards, one H.248 centralized maintenance processor, and N H.248 distributed processors . The combination of these three components can realize large-capacity H.248 protocol processing, and can realize smooth capacity expansion. The external message interface unit of this system is outside the scope of this patent.

The H.248 interface board receives external messages and determines whether to send to the H.248 distributed processor or the H.248 centralized maintenance processor after pre-decoding. The H.248 interface board identifies the processing of each H.248 message entering the system. The device is a key link for realizing H.248 message processing and distribution according to the present invention.

The H.248 centralized maintenance processor receives error messages that are not recognized by the H.248 interface board, or processes maintenance messages for the entire MGC or MGW. The H.248 centralized maintenance processor maintains the normal communication between the MGC and MGW, and maintains them. Consistency of resources and call status. The H.248 distributed processor processes the H.248 messages distributed by the H.248 interface board, and generates call-related The H.248 message is sent to the peer H.248 entity outside the system.

The three components of the H.248 interface board, H.248 distributed processor, and H.248 centralized maintenance processor in this system are responsible for the three processing links of H.248 messages: message pre-decoding and distribution links, and distributed processing. Link and centralized maintenance processing link.

Figure 4 is a flowchart of the interaction between the three main processing links of the H.248 message in the system of Figure 3. The message pre-decoding and distribution link is responsible for the system's external message reception and the pre-decoding judgment of the received message. The centralized maintenance processing link receives the message and pre-decodes the H.248 maintenance message sent from the distribution link, or processes the maintenance request from the distributed processing link. After the processing is completed, the original H.248 distributed processor or the H.248 interface board is notified. Inform the peer H.248 entity. The distributed processing link mainly processes the messages from the message pre-decoding and distribution link, and generates messages according to the call progress and sends them to the peer H.248 entity. The messages between the H.248 distributed processors include multiple actions in a transaction ( Action) to process requests and process response messages.

Figure 5 is a processing flowchart of the pre-decoding and distribution of the H.248 message in Figure 4. As a resource, the transaction ID (transaction ID) of the transaction request sent by the local MGC / MGW is allocated to each H.248 distributed processor. The transaction ID and processor can be found at the message pre-decoding and distribution point. According to the comparison table, when the transaction request sent by the H.248 entity responds to the other party, the H.248 interface board first finds the corresponding processor according to the transaction response ID. For MGC, the H.248 interface board continues to decode the message to the EVENT Request ID. The event number identifier is used as a resource. The H.248 interface board resides for each H.248 distributed processor. Event number identification distribution table. If the received H.248 message has an event number message, the corresponding H.248 distributed processor can be found according to the table. For MGW, the context ID (ContextlD) assigned by MGW is a resource, and a period is allocated for each H.248 distributed processor. The comparison between the context ID and the processor can be found at the message pre-decoding and distribution point. table. If the received message is a request from the peer H.248 entity, the H.248 interface board continues to decode to find the first ContextID. If the context ID is new or empty, specify an H.248 distributed processor The message is processed, and if the context identifier conforms to the context range of a H.248 distributed processor, the message is sent to the H.248 distributed processor for processing. Has an illegal transaction ID or event number ID or The messages identified by the text, or the messages and error messages for the ROOT and "-" flags are processed by the centralized maintenance processor in H.248.

The H.248 centralized maintenance processor maintains state synchronization between H.248 communication entities. When the H.248 protocol distributed processing system is on the MGC, it accepts MGW registration, and when the H.248 protocol distributed processing When the system is on the MGW, it initiates registration with the MGC. This section also completes the processing of error messages and resource availability maintenance messages, such as the H.248 SERVICE CHANGE command.

Figure 6 is the processing flow of the H.248 message distributed processing link in Figure 4. The relationship between the H.248 distributed processors is equal to each other, and there is no belonging relationship. There are two types of H.248 messages processed on the H.248 distributed processor. One is the H.248 message whose context belongs to the local processor, and the other is the context which is not managed by the local processor H.248 message. The processor will process its own context, and then determine whether the message was forwarded by another H.248 distributed processor. If so, it will respond to the forwarding processor with the processing result, and if not, encode it as H.248. The message is sent out. If the processor finds that there is a context that needs to be processed by other processors after processing its own context, it first distributes it to the relevant H.248 distributed processor, and waits for other relevant processors to process the other contexts and encode them into H together. .248 message sent. The secondary distribution of H.248 distributed processors is to deal with the situation where multiple actions occur in a transaction. If the processors corresponding to these contexts are different, the H.248 distributed processor that first receives the H.248 message needs to assume the tasks of distribution, waiting for a response, and then uniformly encoding.

In the model of the system and method of the present invention, the H.248 interface board and the H.248 centralized maintenance processor have a small load, and the H.248 distributed processor processes the final decoding of all messages and a large call load. H can be superimposed as needed .248 distributed processor to increase H.248 processing power. The amount of pre-decoding is small. A complete NGN call is about 20 to 60 H.248 messages. The H.248 centralized maintenance processor can only process at most one of the messages (you can only process registration messages as required). Therefore, in theory, this model The processing capacity can reach more than 20 times of the simple processing method, which can meet all the large-capacity H.248 processing requirements in NGN.

Claims

Claim

1. A H.248 protocol distributed processing system over MGC / MGW, characterized in that the system includes:

One or more H.248 interface boards, each of the H.248 protocol interface boards is configured to receive external H.248 messages, identify and distribute each H.248 message to its corresponding processor;

One or more H.248 distributed processors, each of which is configured to process H.248 messages distributed by the H.248 interface board and generate call-related H.248 The message is sent out;

The H.248 centralized maintenance processor is configured to process an error message sent by each of the H.248 interface boards and a maintenance message for the MGC / MGW, and maintain normal communication between the MGC and the MGW.

2. The system according to claim 1, wherein each of the H.248 interface boards further pre-decodes the received H.248 message.

3. The system according to claim 2, wherein the pre-decoding is incomplete decoding.

4. The system according to claim 1, wherein each of the H.248 distributed processors has a MGC / MGW system resource interval allocated to it.

5. The system according to claim 4, wherein, when the system is on an MGC, a system resource interval of each of the H.248 distributed processors includes a transaction ID interval allocated to the MGC and Event request ID interval; when the system is on the MGW, the system resource interval of each of the H.248 distributed processors includes a transaction ID interval and a context ID interval allocated to the MGW.

6. The system according to claim 5, wherein each of the H.248 interface boards has a comparison table between the MGC / MGW system resource interval and each H.248 distributed processor.

7. The system according to claim 6, wherein each of the H.248 The interface board identifies the H.248 distributed processor corresponding to the received H.248 message according to the comparison table, and distributes the H.248 message to the H.248 distributed processor for processing.

8. A distributed processing method for H.248 protocol using the system according to claim 1, characterized in that the method comprises the following steps:

Each H.248 interface board receives an external H.248 message, identifies the H.248 message, and distributes the H.248 message to a corresponding processor;

Each of the H.248 distributed processors receives and processes the H.248 messages distributed by each of the H.248 interface boards, and generates call-related H.248 messages to send outward;

The H.248 centralized maintenance processor receives and processes error messages sent by each of the H.248 interface boards and maintenance messages for the MGC / MG W.

9. The method according to claim 8, wherein each of the H.248 interface boards has a comparison table between the MGC / MGW system resource interval and each H.248 distributed processor, wherein the identifying and distributing The steps further include:

The H.248 interface board decodes the H.248 message received by the fan; the H.248 interface board determines the processor corresponding to the H.248 message according to the decoding result and the comparison table;

The H.248 interface board sends the H.248 message to the determined processor.

10. The method according to claim 9, wherein the determining step further comprises:

The decoded H.248 message is judged. If the H.248 message is a transaction response message sent by an external entity to the MGC / MGW, the matching table is searched according to the decoded transaction ID to determine the H H.248 distributed processor to which the .248 message belongs;

If the H.248 message is a transaction request from an external entity to the MGC / MGW, when the system is on the MGC, the comparison table is searched according to the decoded event request ID to determine the corresponding H.248 message. H.248 distributed processor; When the system is on the MGW, the lookup table is searched according to the context ID obtained by decoding to determine the corresponding H.248 distributed processor in the range of the context ID. If the context ID is new or empty, an H is specified. A .248 distributed processor, and the H.248 distributed processor assigns a value to the context ID;

If the H.248 message is decoded to obtain an error message and a maintenance message for the MGC / MGW, the H.248 message is sent to the maintenance processor in the H.248 set.

11. The method according to claim 8, further comprising: after each H.248 distributed processor has processed the content belonging to the processor in the H.248 message, determining whether there is still Content that needs to be processed by other processors, if there is, distribute the H.248 message to the relevant processors, and wait for the processing result;

After processing all the contents in the H.248 message, the H.248 distributed processor judges whether the H.248 message is forwarded by other processors, and if so, responds to the source processor with the processing result, If not, the processing result is encoded into an H.248 message and sent.

12. The method according to claim 8, further comprising: the H.248 centralized maintenance processor further receiving and processing a maintenance request message from each of the H.248 distributed processors. And notify the H.248 distributed processor of the completion of the processing.