JP2010074310A - Gateway system and load distribution program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To distribute loads so as not to cause overload, even if the load of a server to which processing is distributed supposedly is increased. <P>SOLUTION: When a calling request of a mobile terminal 60 is received, a server 110A1 selects a server to establish a call from servers 110A1-110A4, including itself. Then, it is selected, on the basis of the loads put on the respective servers at present and the number of calls established by the respective servers at present. When the load are compared, the load applied on the server 110A4 is the lowest and the load put on the server 110A3 is the second lowest. However, the number of calls held by the server 110A4 at present is far larger than the number of calls held by the server 110A3 at present. At the present situation, even though the server 110A4 is in a minimal load state, the load may be increased far exceeding the load put on the server 110A3, depending on the use condition of a user. Here the server 110A1 selects the server 110A3. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention includes a gateway system that distributes processing related to transmission / reception of user data to a plurality of servers by a representative server having a plurality of servers and including a call control unit among the plurality of servers, and a load distribution program that causes the representative server to execute the processing. In particular, the present invention relates to a gateway system and a load distribution program that can perform load distribution so as not to be overloaded even if the load of a server to which processing is distributed increases.

  In recent years, in mobile communication systems, Super 3G has attracted attention as the next system of the third generation, which is currently mainstream. In a super 3G wireless communication system, introduction of a gateway system having a plurality of servers is considered.

  In the gateway system, the representative server determines which server performs processing related to transmission / reception of user data to / from a subordinate mobile terminal. Specifically, the representative server receives a location registration request from the mobile terminal, selects one server from all servers including its own device, and determines the location of the mobile terminal for the selected server. Instruct to register. When the representative server receives a call request from a mobile terminal whose location has been registered, the representative server instructs the server that has performed location registration of this mobile terminal to establish a call. The server that has established a call with the mobile terminal performs processing related to transmission / reception of user data to / from the mobile terminal.

  As described above, when a predetermined process is distributed to a plurality of servers, it is desirable to perform load distribution so that the load applied to each server is equal. Therefore, for example, it is conceivable to apply a technique in which the load states of all servers are acquired at the timing of distributing the processing, the server with the smallest load state is selected, and the selected server is in charge of the processing.

JP-A-11-215180

  However, the above-described conventional technique has a problem that the load of the server to which the load is currently selected as the minimum and the processing is distributed increases to an unacceptable level depending on circumstances. This is because, in the case of Super 3G, the load applied to the server varies depending on the usage status of the user. Therefore, even if the load of the server to which the process is distributed increases, it is an issue to distribute the load so as not to be overloaded.

  The present invention has been made to solve the above-described problems of the prior art, and even if the load of the server to which the process is distributed increases, it is possible to perform load distribution so as not to be overloaded. An object is to provide a gateway system and a load balancing program.

  In order to solve the above-described problems and achieve the object, this system includes a plurality of servers, and a representative server including a call control unit among the plurality of servers performs processing related to transmission / reception of user data. A gateway system for distributing to a server, wherein the representative server is configured such that when the call control unit receives a call request from a mobile terminal, the load currently applied to each server and the potential for load generation on each server And a call processing means for registering the location of the mobile terminal with the server selected by the call time selection means and establishing the call as a server to establish a call. It is a requirement to have

  In addition, this program is a load distribution program that causes a computer to execute a method for distributing processing related to transmission / reception of user data to a plurality of servers, and is currently applied to each server when a call request from a mobile terminal is received. A call selection procedure for selecting a server based on the load and the load generation potential of each server, a server selected by the call selection procedure, registering the location of the mobile terminal, and It is a requirement that the computer execute a call processing procedure determined by the server to be established.

  According to one aspect of the system disclosed herein, even if the load of the server to which the process is distributed increases, the load can be distributed so as not to be overloaded.

  Exemplary embodiments of a gateway system and a load distribution program according to the present invention will be explained below in detail with reference to the accompanying drawings.

  First, an outline of the gateway system according to the present embodiment will be described. The gateway system according to the present embodiment is one node in a wireless communication system compliant with Super 3G. FIG. 1 is a diagram illustrating an example of a wireless communication system compliant with Super 3G. The wireless communication system 200 includes base stations 10A1 to 10An, a hub 20, a hub 30, a router 40, and a gateway system 100, and is connected to a network 50 that is a communication network such as a public telephone network or the Internet. . The hub 20 performs data transfer between the base stations 10A1 to 10An and the gateway system 100, and the hub 30 performs data transfer between the gateway system 100 and the router 40. The router 40 performs data transfer between the wireless communication system 200 and the network 50.

  The radio communication system 200 performs packet processing performed by a base station, an RNC (Radio Network Controller), an SGSN (Serving GPRS Support Node), and a GGSN (Gateway GPRS Support Node) in a radio communication system compliant with the third generation. 10A1-10An and the gateway system 100 are considered. This is because the functions of the RNC and SGSN are distributed to the base stations 10A1 to 10An and the gateway system 100 to centralize functions and improve efficiency.

  As a result, the gateway system 100 is given three functions. In other words, it is called “maintenance” and functions as an interface with the operation terminal, called “MME (Mobile Management Entity)”, and is processing for a location registration request, call connection request, call termination request, etc. from the mobile terminal. A function for performing C-Plane control, called “UPE (User Plane Entity)”, is a function for performing U-Plane control, which is processing related to transmission and reception of user data.

  The gateway system 100 includes servers 110A1 to 110An, all servers have maintenance and UPE functions, and the server 110A1 has MME functions. The server 110A1 having the MME function determines which server performs processing related to transmission / reception of user data to / from mobile terminals under the base stations 10A1 to 10An. In that case, it is desirable to perform load distribution so that the load applied to each server is equal.

  Therefore, as illustrated in FIG. 2, when the server 110A1 receives a call request from the mobile terminal 60 via a base station (not shown), the processing load among the servers 110A1 to 110A4 including its own device. A server with a relatively low and a relatively small number of calls currently established is selected. Here, the mobile terminal 60 actually makes a location registration request before making a call request. However, even if the server 110A1 accepts the location registration request of the mobile terminal 60, the server 110A1 does not finalize the location registration of the mobile terminal 60 by the server, such as suspending the location registration instruction to the server, and puts it in an undecided state. For convenience of explanation, the gateway system 100 includes servers 110A1 to 110A4.

  In FIG. 2, when the load currently applied to each server is compared, the load applied to the server 110A4 is the lowest and the load applied to the server 110A3 is the second lowest. However, the number of calls currently being handled by server 110A4 is much larger than the number of calls currently being handled by server 110A3. In that case, the server 110A4 is currently in a minimum load state, but depending on the usage status of the mobile terminal by each user, the load may increase until the load on the server 110A3 is far exceeded.

  Therefore, the server 110A1 selects the server 110A3, instructs the server 110A3 to register the location of the mobile terminal 60 and to establish a call for the mobile terminal 60. Note that the server 110A1 also compares the current load status of each server even when the server needs to be selected at the time of incoming call to the mobile terminals under the base stations 10A1 to 10An. The server is selected by comparing the number of calls held.

  As described above, the server 110A1 selects a server after comparing not only the current load state of each server but also the number of calls that can determine the load generation potential. By doing so, even if the load of the server to which the process is distributed increases, load distribution can be performed so as not to be overloaded. As long as the gateway system 100 has a plurality of servers and the server distributes processing to each server by a server in charge of call control such as MME, the wireless communication system 200 can use any mobile communication system. It may be compliant.

  Next, the configuration of the server 110A1 will be described with reference to FIG. FIG. 3 is a block diagram showing a configuration of the server 110A1. FIG. 3 shows only the configuration related to the process distribution by the server 110A1.

  As illustrated in FIG. 3, the server 110 </ b> A <b> 1 includes a storage unit 120, a CPU (Central Processing Unit) usage rate information acquisition unit 130, a call control unit 140, and a server selection unit 150. The storage unit 120 stores various types of information, and stores subordinate mobile terminal information 121, call count information 122, and CPU usage rate information 123.

  The subordinate mobile terminal information 121 is an IP (Internet Protocol) address of a mobile terminal whose current location has been registered, and is updated by the call processing unit 141 described later.

  The number-of-calls information 122 is information for comparing the number of calls currently established in each server. As shown in FIG. 4, the server identification information and the number-of-calls information are associated with each other. . The server identification information is information for identifying the servers 110A1 to 110An. The call number information is information indicating the number of calls currently established by the server.

  The CPU usage rate information 123 is information for comparing the load currently applied to each server, and as shown in FIG. 5, the server identification information and the CPU usage rate are associated with each other. The server identification information is information for identifying the servers 110A1 to 110An, similarly to the server identification information of the call count information 122. The CPU usage rate is information indicating how much the CPU of the server is currently used. The information associated with the server identification information may be any information as long as it can compare the load currently applied to each server, for example, the memory usage rate.

  The CPU usage rate information acquisition unit 130 is a processing unit that periodically updates the CPU usage rate information 123. Specifically, the CPU usage rate information acquisition unit 130 accesses the servers 110A1 to 110An at predetermined intervals to acquire the CPU usage rate. Then, the CPU usage rate information acquisition unit 130 updates the CPU usage rate information 123 based on the acquired CPU usage rate.

  The call control unit 140 is a processing unit that performs processing related to the function of the MME, and includes a call processing unit 141 and an incoming call processing unit 142. The outgoing call processing unit 141 performs processing related to the outgoing call of the mobile terminal, and the incoming call processing unit 142 performs processing related to the incoming call to the mobile terminal. In processing by each processing unit, it is necessary to select a server in order to distribute processing related to transmission / reception of user data when a subordinate mobile terminal makes a call or when a mobile terminal receives a call. In that case, each processing unit instructs the server selection unit 150 to select a server.

  When the call processing unit 141 receives a call request from the mobile terminal via the base station, the call processing unit 141 instructs the server selection unit 150 (to be described later) to select a server. Actually, the call processing unit 141 receives a location registration request before receiving a call request from a mobile terminal. However, the call processing unit 141, for example, only changes the subscriber state (changes from the LTE (Long Term Evolution) _DETACHED state to the LTE_IDLE state) and instructs the server to perform location registration. Hold.

  After instructing the selection of the server, the calling processing unit 141 designates the server from the server selecting unit 150. Therefore, the calling processing unit 141 instructs the designated server to perform location registration, and displays the subordinate mobile terminal information 121. Update and add the IP address of the mobile terminal that made the location registration request. In addition, the call processing unit 141 updates the call number information 122 to increase the number of calls of the designated server by one. Then, the call processing unit 141 instructs the designated server to establish a call. In addition, when the call request of the mobile terminal is made after requesting location registration different from the previous time, the call processing unit 141 deletes the location registration to the server that originally registered the location. To instruct.

  In addition, when receiving a call termination request from a mobile terminal, the call processing unit 141 instructs a server that has established a call with the mobile terminal to terminate the call. Then, the call processing unit 141 updates the call number information 122 and decreases the number of calls of the server that has instructed termination by one.

  The incoming call processing unit 142 receives an ARP (Address Resolution Protocol) request packet from the router 40 in FIG. As a stage until an ARP request packet is transmitted, the router 40 receives an IP packet from a node in the network 50 toward a mobile terminal in the wireless communication system 200. Then, the router 40 transmits an ARP request packet conforming to the ARP as a broadcast IP packet in order to obtain a frame address (MAC (Media Access Control) address) from the destination IP address included in the packet.

  The incoming call processing unit 142 compares the destination IP address included in the received ARP request packet with all IP addresses included in the subordinate mobile terminal information 121. If there is an IP address that matches the destination IP address, the incoming call processing unit 142 instructs the server selection unit 150 to select a server. On the other hand, the incoming call processing unit 142 discards the ARP request when the IP addresses do not match. As a result, the router 40 discards the received IP packet in accordance with the ARP, with an ARP response reception timeout. Here, the other server similarly receives the ARP request packet from the router 40, but the server 110A1 causes the other server to discard the ARP request unconditionally and acts as an ARP response.

  After instructing the selection of the server, the incoming call processing unit 142 designates the server from the server selecting unit 150, and therefore instructs the designated server to perform location registration. Then, the incoming call processing unit 142 updates the call number information 122 and increases the number of calls of the designated server by one. The incoming call processing unit 142 also registers the original location if the server that originally registered the location of the mobile terminal having the IP address that matches the destination IP address is different from the server specified by the server selection unit 150. Instruct the server that had been deleted to delete the location registration. Then, the incoming call processing unit 142 acquires the MAC address of the designated server from the MAC address data of all the servers stored in advance in its own device, sets it as the destination MAC address in the ARP response packet, and the router 40 An ARP response packet is transmitted to.

  Thereafter, the router 40 sets the destination MAC address included in the ARP response packet as the destination of the IP packet, and transmits the IP packet. Then, the server selected by the server selection unit 150 receives the IP packet from the router 40 and notifies the server 110A1 of the incoming call. The incoming call processing unit 142 receives the incoming call notification and performs paging through the base stations 10A1 to 10An.

  When instructed by the call control unit 140 to select a server, the server selection unit 150 selects a server based on the call count information 122 and the CPU usage rate information 123. In the case of the super 3G, the number of calls established between the mobile terminal and the server and the high load on the server are not necessarily proportional. For example, even in a server that has established more calls than other servers, the load state may be minimized if the call is stable. However, such a server may suddenly increase its load and become much higher than other servers. On the other hand, even in a server that has established fewer calls than other servers, the load state may be maximized depending on the usage status of the user. However, such a server may suddenly go down and become much lighter than other servers.

  Therefore, for example, for each server, the server selection unit 150 multiplies the number of calls currently established by the CPU unused rate, and obtains a second call number obtained by further multiplying the multiplication result by a predetermined coefficient. And the server selection part 150 calculates | requires the ratio of the 2nd call number in the number of calls currently established by a percentage, and adds it to CPU usage rate, respectively, calculates an addition value for every server, and calculated addition Select the server with the lowest value. As described above, the server selection unit 150 selects a server by adding the number of calls to the CPU usage rate as well as the CPU usage rate. For example, if the selection of the server takes into account the potential for occurrence of load in addition to the current load, such as multiplying the calculated percentage by the CPU usage rate and selecting the server with the lowest calculated multiplication value, The server may be selected with any logic. In addition to the number of calls, the information that is added to the current load when selecting a server may be the data transfer amount of the current server. Also, the calculation may be performed using the call bandwidth instead of the number of calls. As described above, the server selection unit 150 selects a server in consideration of not only the current load state of each server but also the number of calls that can determine the potential of load generation in the future. By doing so, even if the load of the server to which the process is distributed increases, load distribution can be performed so as not to be overloaded.

  The server selection unit 150 refers to the CPU usage rate information 123 that is updated at a predetermined interval by the CPU usage rate information acquisition unit 130 in order to compare the current load. Any method can be used as long as the loads can be compared. For example, the servers 110A1 to 110An may be accessed by themselves and the CPU usage rate may be acquired.

  Next, the processing operation related to the transmission of the mobile terminal by the server 110A1 will be described with reference to FIG. FIG. 6 is a flowchart showing a flow of processing relating to the transmission of the mobile terminal by the server 110A1.

  As shown in FIG. 6, in the server 110A1, when the call processing unit 141 receives a call request (Yes at Step S101), the server selection unit 150 receives instructions from the call processing unit 141 and receives all instructions. The CPU usage rate information 123 is referred to (step S102), and the call count information 122 is referred to (step S103). Then, the server selection unit 150 selects a server based on the CPU usage rate information 123 and the call number information 122 (step S104), and the call processing unit 141 performs location registration with the server selected by the server selection unit 150. (Step S105). Then, the call processing unit 141 updates the call number information 122 (step S106), instructs the server selected by the server selection unit 150 to establish a call (step S107), and ends the process. Note that the update processing of the call number information 122 shown as step S106 may be performed at any timing as long as the server is selected. For example, the call processing unit 141 may update the call number information 122 after instructing to establish a call.

  Next, the processing operation related to the incoming call of the mobile terminal by the server 110A1 will be described with reference to FIG. FIG. 7 is a flowchart showing the flow of processing related to the arrival of a mobile terminal by the server 110A1.

  As shown in FIG. 7, in server 110A1, when receiving an ARP request packet (Yes at step S201), incoming call processing unit 142 compares the destination IP address in the packet with the IP address of the mobile terminal whose current location has been registered. (Step S202). If there is an IP address that matches the destination IP address as a result of the comparison (Yes at step S203), the server selection unit 150 refers to the CPU usage information 123 of all servers in response to an instruction from the call processing unit 141. (Step S204), the call number information 122 is referred to (Step S205). On the other hand, when there is no IP address that matches the destination IP address (No at Step S203), the incoming call processing unit 142 discards the ARP request (Step S210).

  The server selection unit 150 referring to each information selects a server based on the CPU usage rate information 123 and the call number information 122 (step S206), and the incoming call processing unit 142 is located at the server selected by the server selection unit 150. An instruction is given to perform registration (step S207). Then, the incoming call processing unit 142 updates the call number information 122 (step S208), sets the MAC address of the server selected by the server selection unit 150 in the ARP response packet, and transmits it (step S209). To do. Note that the update processing of the call number information 122 shown as step S208 may be performed at any timing as long as the server is selected. For example, the incoming call processing unit 142 may update the call number information 122 after transmitting the ARP response packet.

  As described above, in the gateway system 100 according to the present embodiment, the server 110A1 selects a server in consideration of not only the current load state of each server but also the number of calls that can determine the load generation potential. . By doing so, even if the load of the server to which the process is distributed increases, load distribution can be performed so as not to be overloaded.

  The configuration of the server 110A1 according to the present embodiment shown in FIG. 3 can be variously changed without departing from the gist of the present invention. For example, the functions of the CPU usage rate acquisition unit 130, the call control unit 140, and the server selection unit 150 of the server 110A1 can be implemented as software, and can be executed by the CPU to realize functions equivalent to the server 110A1. An example of a computer that executes a load distribution program 341 in which the functions of the respective units are implemented as software is shown below.

  FIG. 8 is a functional block diagram showing a computer 300 that executes the load distribution program 341. The computer 300 includes a CPU 310 that executes various arithmetic processes, a network interface device 320 that exchanges data with other computers via a network, and a RAM (Random Access Memory) 330 that temporarily stores various information. ROM (Read Only Memory) 340 is connected by a bus 350.

  The ROM 340 stores a load distribution program 341 having the same functions as those of the CPU usage rate acquisition unit 130, the call control unit 140, and the server selection unit 150 illustrated in FIG. 3, and the call number information 122 illustrated in FIG. Further, call number data 342 and CPU usage rate data 343 corresponding to the CPU usage rate information 123 are stored.

  When the CPU 310 reads the load distribution program 341 from the ROM 340 and develops it in the RAM 330, the load distribution program 341 functions as the load distribution process 331. Then, the load distribution process 331 expands the information read from the call count data 342 and the CPU usage rate data 343 to an area allocated to itself on the RAM 330 as appropriate, and performs various data processing based on the expanded data and the like. Execute.

  Note that the load distribution program 341 is not necessarily stored in the ROM 340, and the computer 300 may read and execute the program stored in a storage medium such as a memory card. Further, the computer 300 stores the program in another computer (or server) connected to the computer 300 via a public line, the Internet, a LAN (Local Area Network), a WAN (Wide Area Network), or the like. You may make it read and run a program from these.

It is a figure which shows an example of the radio | wireless communications system based on super 3G. 1 is a diagram for explaining an overview of a gateway system 100. FIG. It is a block diagram which shows the structure of server 110A1. It is a figure which shows the specific example of the call number information. It is a figure which shows the specific example of CPU utilization rate information. It is a flowchart which shows the flow of the process regarding the time of transmission of the mobile terminal by server 110A1. It is a flowchart which shows the flow of the process regarding the time of the incoming call of the mobile terminal by server 110A1. 2 is a diagram illustrating a computer 300 that executes a load distribution program 341. FIG.

Explanation of symbols

10A1 to 10An Base station 20, 30 Hub 40 Router 50 Network 60 Mobile terminal 100 Gateway system 110A1 to 110An Server 120 Storage unit 121 Subordinate mobile terminal information 122 Call count information 123 CPU usage rate information 130 CPU usage rate information acquisition unit 140 Call control 141 Outgoing call processing unit 142 Incoming call processing unit 150 Server selection unit

Claims (8)

A gateway system that has a plurality of servers and distributes processing related to transmission / reception of user data to the plurality of servers by a representative server having a call control unit among the plurality of servers,
The representative server is
When the call control unit receives a call request from a mobile terminal, a call time selection means for selecting a server based on the load currently applied to each server and the load generation potential of each server; ,
Call processing means for registering the location of the mobile terminal as the server selected by the call time selection means and establishing the call as a server;
A gateway system comprising:   2. The gateway system according to claim 1, wherein the call processing unit determines the load generation potential based on the number of calls currently established by each server.   2. The gateway system according to claim 1, wherein the call processing unit determines the load generation potential based on a current data transfer amount of each server. The representative server is
Based on the load currently applied to each server and the load generation potential of each server when the call control unit receives an inquiry about the transmission destination of data in which the subordinate mobile terminal is set as the destination An incoming call selection means for selecting a server;
An incoming call processing means for determining the server selected by the incoming call selection means as a server for registering the location of the mobile terminal, and for confirming as a transmission destination that responds to the inquiry source of the transmission destination; The gateway system according to claim 1, 2, or 3.   5. The gateway system according to claim 4, wherein the incoming call processing means determines the load generation potential based on the number of calls currently established by each server.   The gateway system according to claim 4, wherein the incoming call processing unit determines the load generation potential based on a current data transfer amount of each server. A load distribution program for causing a computer to execute a method for distributing processing related to transmission / reception of user data to a plurality of servers,
A call selection procedure for selecting a server based on the load currently applied to each server and the load generation potential of each server when receiving a call request from a mobile terminal;
A load distribution program that causes a computer to execute a call processing procedure for registering a location of the mobile terminal for a server selected by the call selection procedure and establishing a call to a server. When an incoming call is made to select a server based on the load currently applied to each server and the potential load on each server when inquired about the destination of the data set by the mobile terminal under the destination Selection procedure,
The server selected by the selection procedure at the time of the incoming call is determined as a server for registering the location of the mobile terminal, and an incoming call processing procedure for determining as a transmission destination responding to the inquiry source of the transmission destination is further added to the computer. The load distribution program according to claim 7, wherein the load distribution program is executed.

Images