WO2008061401A1

WO2008061401A1 - A mobile service switching center server of realizing route selecting function

Info

Publication number: WO2008061401A1
Application number: PCT/CN2006/003670
Authority: WO
Inventors: Bo Yang; Yazhou Kong; Jin Wang; Yanxia Hong
Original assignee: Zte Corporation
Priority date: 2006-11-23
Filing date: 2006-12-29
Publication date: 2008-05-29
Also published as: CN101193428A; RU2009124400A; RU2423019C2

Abstract

A mobile service switching center server of realizing route selecting function includes call control module, route selection module, memory module, wherein the call control module receives a call of incoming side gateway and controls MGW on a bearer path to establish bearer according to bearer link information; the route selection module calculates busy degree parameter of each path according to resource configuration amount and resource amount having occupied, selects bearer path and transmits the information to the call control module; the memory module stores configuration parameter of route resource. The present invention conquers the problem that in a multi-gateway model unfair resource allocation between gateways easily leads to the congestion, and enables the whole network to controllably, stably and fairly work.

Description

TECHNICAL FIELD The present invention relates to a mobile service switching center server of a 3GW CDMA-R4 core network in the communication field, and more particularly to a mobile service switching center server implementing a routing function. Background technique

3GWCDMA-R4 (3G refers to the third generation mobile communication technology, WCDMA is the wideband code division multiple access, and R4 is the version number) The basic architecture of the network is the separation of bearer and control, and the control part of the call is concentrated in the MSC Server (Mobile Service Switching Center) On the MGW (media gateway), the typical network architecture is that one MSC Server controls one MGW, and multiple sets of such devices are interconnected to form a network. . However, the disadvantage of such a networking mode is that each MGW needs an MSC Server to control it, which will result in repeated investment of the MSC Server.

In the development process of 3G network, in order to save cost, a network architecture model is evolved. Two, three or more MGWs are simultaneously registered to one MSC Server. This network model is called a multi-gateway model. Because this network model is practical and economical, it is quickly applied. With the large-scale application of this model, more and more large-area multi-gateway models will be used to network, but in large areas. Inter-office interconnection between MSC Servers is used. After this application reaches a certain level, there will be more and more gateways under the same MSC Server, and the traffic between these gateways will be larger and larger, and the number of resources among multiple gateways is limited. Therefore, it is possible that some gateways are busy, and some gateways are not too busy. How to balance the load of these traffic traffic to other gateways, and how to effectively avoid the exhaustion of resources between some gateways when the traffic suddenly increases, while other gateways are very idle, it is necessary solved problem. At present, the multi-gateway model has just started, so the traffic between the gateways is not very large. The resource allocation between the gateways is allocated according to the shortest route. However, this method of routing is simple, and it is not able to make good use of resources between gateways. Suppose there are two gateways. There are two paths between them. One is short and the other is long. When there is a lot of traffic between the two gateways, the shortest path between the two gateways will be very busy. . When the traffic volume increases to a certain extent, the resources on the shortest path will be exhausted, and other calls that wish to pass this path will be rejected; and the longer path is idle because there is no traffic. Then appear on a shorter path Congested. If you can share a part of the traffic to take the longer path before the resources on the shortest path are exhausted, then there will be no traffic on the longer route, but the resources on the shorter path are exhausted. Happening. Therefore, there is another way of routing, which is routed according to the load sharing method. This method simply assigns traffic to two different paths according to the proportion of resources as the weight, and the resource allocation is relatively large. The amount of traffic allocated is larger, and the amount of traffic allocated less is less. However, this method does not consider the user's talk time. Some users may hang up soon after making a call, and some users may The call is long and the bearer resources are always occupied. On the other hand, the current call path may not only have traffic between the two gateways, but there may be traffic between other gateways, and the two gateways do not have busy traffic. The degree of change is the basis for routing, so the final result will be disordered and uncontrollable, and it is also prone to resource exhaustion, but the probability of this situation is reduced. It can be seen that the technology and routing strategy currently used in practical applications are actually unfair, and can not better coordinate the resource utilization and allocation between gateways in the multi-gateway model. SUMMARY OF THE INVENTION The technical problem to be solved by the present invention is to provide a mobile service switching center server that implements a routing function, which overcomes the unfair resource allocation in the multi-gateway model and easily causes congestion and call loss, thereby making the entire network controllable. Run stably and fairly. In order to solve the above technical problem, the present invention provides a mobile service switching center server that implements a routing function. The server MSC Server governs multiple media gateways MGW, and is characterized by: including a call control module, routing The module and the storage module, wherein the call control module is configured to receive the call of the ingress side gateway Ni, and transmit the information of the ingress side gateway Ni and the outgoing side gateway Nj to the routing control module, and accept the bearer path information returned by the routing control module. The routing control module is configured to select a path between the ingress side gateway Ni and the outgoing side gateway Nj according to the network topology configuration, and each path is selected according to the network topology configuration. Recording in the form of a line set, calculating the tension degree parameter of each path according to the parameters of the network management configuration and the total number of line resource configurations in the line set stored in the storage module and the total number of resources already occupied, and selecting the path with the lowest degree of tension as the carrying path. Transmitting the information to the call control module; The storage module is configured to store configuration parameters and occupancy parameters of the routing resource, and parameters of the network management configuration. Further, the line set recorded by the routing control module is a line set whose hop count is less than or equal to the specified fair hop count hop. Further, the routing control module separately calculates a corrected line tension factor = f (line tension factor) for each line in the line set on each path, wherein the line tension factor = the number of occupied resources / the total number of resource configurations, The domain of the function f ( X ) is: 0<=χ<=1; and f ( X ) is a smooth function with f(0)=0 and f(l)=infinity; then the correction of each path is calculated separately. The average line tension degree, the corrected average line tension is the sum of the corrected line tension factors of each line in the line set, and divided by the hop count of the path, and the path with the smallest corrected average line strain is selected as the bearer among all the paths. path. Further, the routing control module separately calculates a line tension factor of each line in each line in each path. When the tension factor of each line of the shortest path is less than a specified load starting threshold, the shortest path is selected as the shortest path. The bearer path of the current call; otherwise, the bearer path is selected between multiple gateways according to the parameters configured by the network management system and the total number of resource configurations stored in the storage module and the total number of resources already occupied. Further, the load sharing 4 positive function f ( X ) = l / (lx) - l. Further, the routing module further includes a load sharing control sub-module, configured to calculate a load sharing effect according to a load sharing correction function of the network management, and further load control results of the bearer link between the gateways and current resource occupancy parameters. Dynamic statistics are reported to the NMS or storage module. The invention has the following beneficial effects:

1. Compared with the uniform or weighted routing method according to the traffic load sharing, the method of balancing or weighting according to the traffic load sharing only considers the current traffic arrival rate, and does not consider different The difference in the length of the telephone conversation time is therefore prone to large fluctuations in fairness. However, the present invention does not use the incoming traffic as the object of load control, but uses the current total resource allocation and the total number of resources already occupied as an important factor for evaluating different routes. It utilizes all MGWs. It is within the jurisdiction of the same MSC Server, and MSG Server can easily observe the advantages of the busyness of all nodes in the storage module to control the uniform load sharing on different paths. 2. The current 3 GW CDMA-R4 core network multi-gateway model with no load sharing function selects the shortest path as the current path, so that when the traffic between two gateways is relatively large, a jump occurs. The traffic on the shortest path with a small number is extremely busy, even the resources are exhausted, and the call loss occurs. However, the path with a long hop count has no traffic. In fact, as long as the hop count of one path is within the tolerable range. You should share some of the traffic to these paths to ease the pressure on shorter paths with fewer hops. Therefore, the present invention limits the range of routing to a fair hop count that can be configured by a network management system, so that the traffic can be shared to a longer path, and the user can be restricted on these longer paths to avoid The path is too long and the overhead is too large.

3. Add up the busyness factors of all the lines on a path and average them. By using this method, when one line on this path is busy and the other line is idle, the average will cover the line. The busy part of the resource is tight, and this time, as a route to be selected, may increase the tension of the line where the line is tight, which in turn causes congestion. After the correction is performed by the function f(x), the final corrected average line tension becomes sensitive to congestion, and it can be reflected in the correction of the average line tension immediately when congestion is about to occur. The value, and thus the probability that this path is selected, will drop rapidly, thus playing a role in avoiding congestion. When a call on a line that is about to be congested is released, the chance that the path may be selected increases. Therefore, the invention can effectively avoid congestion.

4. The present invention sets a load start threshold, taking into account both the shortest route and the congestion avoidance. When the line tension factor of each line on the shortest path is less than the load control start threshold, the user is indicated. It is considered that congestion will not occur at this time, so the shortest path should be selected as the path taken by the bearer of the current call. When the line tension factor of any line on the shortest path is greater than the congestion control threshold, a load sharing mechanism needs to be started to help the shortest path to share the traffic by other paths. Therefore, the present invention takes into consideration the advantages of both the shortest route and the congestion avoidance. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of a connection between an MSC Server and an MGW according to the present invention; FIG. 2 is a block diagram showing the structure of the MSC Server of the present invention and a communication between the network management system, the MSC Server, and the MGW; FIG. 3 is a flowchart of a method for implementing a routing function according to the present invention; 4 is another flow chart of a method for implementing a routing function according to the present invention; 5 is a diagram showing an example of resource occupation/configuration between gateways in FIG. 1 of the present invention; FIG. 6 is a system for implementing a routing function in a multi-gateway model according to the present invention, in a system of 1 - 6 -> 5 and 1 - 7 — > 5 Resource occupancy/configuration on both paths. DETAILED DESCRIPTION OF THE INVENTION The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. As shown in FIG. 1, node 8 in the figure is an MSG Server, and nodes 1 to 7 are MGWs that are managed by the node 8. The dotted line between node 8 and node from 1 to 7 is the signaling link; the solid line between nodes 1 to 7 is the carrier link. An MSC Server has at least four MGWs, and MGWs form a mesh connection. There are more than one path between two non-adjacent MGWs. As shown in Figure 2, the NMS is used to manage the MSC Server and the MGW, configure parameters and store configuration parameters, and accept and dynamically display resource configuration parameters and resource occupancy parameters between the current gateways transmitted by the MSC Server. The MSC Server of the present invention includes a call control module, a routing module, and a storage module. The call control module is configured to accept a call of the MGW (the MGW is called an ingress side gateway), and determine that the ingress side gateway needs to be determined according to the configuration of the network management. Which MGW is transmitted to the MGW (the MGW is called the outgoing side gateway), and the information is transmitted to the routing control module, and the bearer path information returned by the routing control module is accepted, and the path is controlled according to the bearer path information. All the MGWs establish a bearer; the routing module is configured to select a path between the ingress side gateway and the egress side gateway according to the network topology configuration. Each path is recorded in the form of a line set, and the bearer path is selected between the multiple gateways according to the parameters of the network management configuration and the total number of resource configurations stored in the storage module and the total number of resources already occupied, and the information is transmitted to the call control module. The storage module is configured to store configuration parameters and occupancy parameters of the routing resource, and parameters of the network management configuration; the configuration parameters of the network management include a fair hop count and a load sharing correction function, and the network management system can allocate resources and resources according to the current gateway. Occupancy, dynamic modification and configuration of fair hops. The configuration parameter of the network management system may further include a startup load control threshold for selecting a bearer path.

The MSC Server may further include a load sharing control submodule (not shown;) A sub-module of the routing module can be physically present inside or outside the MSC Server as a separate module or device. The load sharing control sub-module can calculate the load sharing effect according to the load sharing correction function of the network management system, and then report the load control result of the bearer link between the four gateways and the current resource occupancy parameter to the network management system or the storage module. As a basis for users to evaluate the effect of such load control.

The load control of the bearer link on different paths in each gateway by the MSC Server does not affect the traffic and traffic of other MSC servers interconnected with the MSC Server. As shown in FIG. 3, the method for implementing the routing function of the present invention includes the following steps: Step 301, start; Step 302, the call control module receives an incoming side gateway call, and determines an outgoing side gateway according to the information configured by the network management; 303. The call control module sends the ingress side gateway (Ni) and the egress side gateway (Nj) as parameters to the route control module. Step 304: The route control module configures (hop) according to the network topology configuration and the fair hop number (hop) in Ni and Nj. Select all paths with hops less than or equal to hop; each path may consist of one or more lines, and record each path in the form of a line set, assuming a path between Ni and Nj Only one node Nk is passed, then the line set of the path is {(Ni, Nk), (Nk, Nj)}; Step 305, the routing control module reads the storage module according to the line set of each path in step 304. The total resource allocation of these lines (T(Ni, Nk), T(Nk, Nj)) and the number of occupied resources (A(Ni, Nk), A(Nk, Nj)), for the case where the gateway is TDM , T(Ni, Nk) indicates the available configuration between gateways The number of CIC (TDM circuit identification code), A(Ni, Nk) indicates the number of CICs already occupied; For the IP bearer T(Ni, Nk), the total bandwidth of the available bearers between the gateways, A(Ni, Nk) indicates that The bandwidth used is used; Step 306, the line tension factor of the path (Ni, Nk) is calculated according to the formula A(Ni, Nk)/T(Ni, Nk), due to the current occupied resource A(Ni, Nk) total Is less than the configured resource T(Ni, Nk), so A(Ni, Nk)/T(Ni, Nk)<=l; according to the formula, each line in the line set on each path in step 304 is calculated separately. Line tension factor Step 307: Calculate the engineering corrected average line tension of each path separately:

{ A(Ni,Nk)/T(Ni,Nk) +. · .+ A(Nk,Nj)/T(Nk,Nj) }/h = average line tension (Formula 1)

{ f(A(Ni,Nk)/T(Ni,Nk)) +... +f(A(Nk,Nj)/T(Nk,Nj)) }/h=Correct the average line tension (Equation 2) Where ((Ni, Nk), ... (Nk, Nj)) 6 line sets, h is the hop count of the line set; where the function f ( X ) should have the following characteristics: The domain of f ( x ) is : 0<=χ<=1; and if f(X) is a smoothing function with f(0)=0 and f(l)=infinity, then f(A(i,Nk)/T(Ni can be made) , Nk)) is the busiest of 1, making f(A(N, Nk)/T(Ni, Nk)) the least busy for 0. f(x)=l/(l-x)-l has the above condition, and f(x) is substituted into the formula 2 to obtain the formula 3.

{ A(Ni,Nk)/(T(Ni,Nk)- A(Ni,Nk)) +...+ A(Nk,Nj)/(T(Nk,Nj)- A(Nk,Nj)) }/h = Engineering 平均 average line tension (Equation 3) Step 308: Using the engineering corrected average line tension of each path calculated in step 307 to select the engineering correction average line tension is the smallest among all the paths. That path is the path of the current call. Step 309: The routing control module sends the path information to the call control module, and the call control module controls all the MGWs on the path to be established according to the information. Step 310: End. The engineering correction algorithm pointed out in Equation 3 is obtained by substituting the function f(x)=l/(lx)-l into Equation 2, and there are many smoothing functions that conform to the f(x) feature, such as 1/(1). -sin ((兀/2)*x) ) -1 is a function commonly used in higher mathematics, which is consistent with the characteristics of f(x). These smoothing functions should not be limited to f(x)=l/(lx) -l , so the function should be configurable, and the user can measure the impact of different f(x) on the load control effect according to the resource occupancy parameters reported by the routing control module to the network management, and then can be positive f(x) . As shown in FIG. 4, another method for implementing the routing function of the present invention includes the following steps: Step 401: Start; Step 402: The call control module receives an inbound side gateway call, and #_ determines the egress side gateway according to the information configured by the network management. Step 403: The call control module sends the ingress side gateway (Ni) and the outgoing side gateway (Nj) as parameters a routing control module; Step 404, the routing control module selects all paths with a hop count less than or equal to hop between Ni and Nj according to a network topology configuration and a fair hop configuration (hop); each path may be one or more The line consists of each line recorded in the form of a line set. Let a path between Ni and Nj pass through only one node Nk, then the line set of this path is {(Ni, Nk), ( Nk, Nj)}; Step 405: The routing control module reads the total resource configuration (T(Ni, Nk), T(Nk, Nj)) of the lines in the storage module according to the line set of each path in step 404. Number of occupied resources (A(Ni, Nk), A(Nk, Nj)); Step 406, calculating the line tension of the line (Ni, Nk) according to the formula A(Ni, Nk)/T(Ni, Nk) Factor, since the currently occupied resource A(Ni, Nk) is always less than the configured resource T(Ni, Nk), A(Ni, Nk) / T (Ni, Nk) <= l; according to the formula, calculate the line tension factor of each line in each line in each path in step 404; Step 407, the tension level factor of the shortest path line and the network management Comparing the set start load control threshold, if the tension factor of each line of the shortest path is less than the load start threshold, then the shortest path is selected as the bearer path of the current call; Step 408, the route control module sends the path information To the call control module, the call control module controls all the MGWs on the path to establish 7 loads according to the information; step 409, the end. Step 410: If the tension factor of each line of the shortest path in step 407 is greater than the load start threshold, this indicates that the current shortest path is relatively congested, so the load sharing mechanism should be started to calculate the engineering correction of each path separately. The average line tension degree; Step 411, using the calculated positive average line tension degree of each path calculated in the step gland 410, selecting the path with the most corrected average line tension degree as the current call path among all the paths; Step 412: The routing control module sends the path information to the call control module, and the call control module sends the information to the ingress side gateway, and then proceeds to step 409 to end the process. Similarly, in this example, the function used to correct the average line strain is f(x = l/(lx)- l , and there are many smoothing functions that conform to the f(x) feature. These smoothing functions should not be limited to f(x). ) = l / (l - x) - l , so the function should be configurable, and the user can measure the load control effect of different f(x) according to the resource occupancy parameters reported by the routing control module to the network management system. The effect can be positive f(x). Figure 5 is a diagram of the resource occupancy/configuration between gateways of the present invention. The scores among all nodes in the figure, the numerator indicates the total number of resources A(Ni, Nk) that have been occupied between nodes. , and the denominator represents the total number of resources T(Ni, Nk) configured between nodes. This score represents the line tension factor A(Ni, Nk)/T(Ni between two nodes (Ni, Nk). , Nk). Figure 6 is an example of the resource occupancy/configuration on the two paths of 1 - 6 -> 5 and 1 - 7 - > 5. If you follow the formula 1 in these two Select an idle path between the paths, then the average line tension of path 1 > 6 -> 5 is: (0.53 + 0.97 ) /2 = 0.75, the average line tension of the path 1 > 7 - > 5 is: ( 0.85 + 0.85 ) /2 = 0.85. If you directly select the path with less average line tension as the current path, you should choose 1_>6-> 5 is the current path, but it is obvious that at this time, the line utilization rate of the line 6-> 6 between the path 1 -> 6 -> 5 has reached 97%, this time if still at 1 —> 6—〉 5 This route is routed above, and congestion will occur between lines 6–> 5, so you should actually select 1_7–>5 as the current path. If you calculate according to formula 3 If an idle path is selected between the two paths, the corrected average line tension of the path 1 -> 6 -> 5 is: ( 53 / ( 100 - 53 ) + 97 / ( 100 - 97 ) ) / 2 = 33.427, The average line tension of the road 1—>7_>5 is: ( 85/( 100-85 )+85 ( 100-85 ) ) /2=5.56. Select the path with the smaller corrected average line tension as the current path. You should choose 1 > 7 - > 5 as the current path, this time to avoid the path that is about to be congested, this is Explain that after the correction, Equation 3, when any of the lines in a path is about to become congested, the corrected average line tension will become very large, which reduces the possibility of traffic selection on this path. The purpose of congestion avoidance is to make the MSC of the 3GW CDMA-R4 core network in the multi-gateway model by calculating the average line tension degree, setting the load sharing start threshold, and setting the load sharing fair hop count. The routing function that Server has. This function takes into account the advantages of the shortest route and congestion avoidance, effectively coordinating the traffic load between multiple gateways under the same MSC Server, and making the entire network run manually, stably, and fairly.

Claims

A mobile service switching center server implementing a routing function, the server MSC Server administers a plurality of media gateways MGW, and is characterized by: a call control module, a routing selection module, and a storage module,

The call control module is configured to receive the call of the ingress side gateway Ni, and transmit the information of the ingress side gateway Ni and the outgoing side gateway Nj to the routing control module, and accept the bearer path information right returned by the routing control module, according to the information. Controlling the MGW on the bearer path to establish a bearer; the routing module is configured to select a path between the ingress side gateway Ni and the outgoing side gateway Nj according to the network topology configuration, and record each path in the form of a line set. Calculate the tension degree parameter of each path according to the parameters of the network management configuration and the total number of line resource configurations in the line set stored in the storage module and the total number of resources already occupied, and select the path with the lowest degree of tension as the bearer path, and send the information to the call. Control module

The storage module is configured to store configuration parameters and occupancy parameters of the routing resource, and parameters of the network management configuration. The mobile service switching center server implementing the routing function according to claim 1, wherein: the line set recorded by the routing control module is a line set with a hop count less than or equal to a specified fair hop count and hop. The mobile service switching center server implementing the routing function according to claim 2, wherein: the routing control module separately calculates a correction line tension factor of each line in each line on each path = f (line tension) Degree factor), where the line tension factor = the number of occupied resources / the total number of resource configurations, the domain of the function f ( ) is: 0 <= χ < = 1; and f ( X ) is a f (0) = 0, f(l) = infinity smoothing function; calculate the corrected average line tension of each path separately, and correct the average line tension to the sum of the corrected line tension factors of each line in the line set, and divide by the path jump Number, the path with the smallest corrected average line strain is selected among all the paths as the bearer path. The mobile service switching center server implementing the routing function according to claim 2, wherein: the routing control module separately calculates a line tension factor of each line in the line set on each path, when each of the shortest paths The tension factor of a line is less than the specified load start threshold, and the shortest path is selected as the bearer path of the current call; Otherwise, the bearer path is selected between multiple gateways according to the parameters configured by the network management system and the total number of resource configurations stored in the storage module and the total number of resources already occupied. The mobile services switching center server implementing the routing function according to claim 3 or 4, wherein: said load sharing correction function f ( X ) = l / (l - x) - l. The mobile service switching center server implementing the routing function according to claim 3 or 4, wherein the routing module further comprises a load sharing control sub-module, configured to calculate a load sharing effect according to a load sharing correction function of the network management system, Then, the load control result of the bearer link between the gateways and the current resource occupancy parameter are dynamically reported to the network management system or the storage module.