JP5677524B2 - COMMUNICATION CONTROL DEVICE, COMMUNICATION CONTROL SYSTEM, AND COMMUNICATION CONTROL METHOD - Google Patents

Description

  The present invention relates to a communication control device, a communication control system, and a communication control method for controlling congestion.

  Conventionally, as a countermeasure against a network failure, the number of successful access times is counted based on whether there is a response corresponding to the request, the access success rate is calculated from the number of successful access times, and based on the successful access rate, A method for determining whether or not there is a network failure is known (see, for example, Patent Document 1).

JP 2009-171009 A

  By the way, when a failure occurs in the server, the response from the server corresponding to the request from the terminal is delayed. When the terminal that has been communicating with the server fails to receive a response from the server, the terminal repeats retransmission of the request to the server. Then, compared with the case where no failure has occurred in the server, the number of requests transmitted from the terminal to the server increases and congestion occurs.

  However, in the conventional method of determining whether there is a network failure based on the access success rate obtained from the number of successful access attempts, the number of successful access times is counted based on whether there is a response corresponding to the request. As a result, the calculation of the number of successful accesses is delayed. For this reason, there has been a problem that congestion cannot be detected at an early stage.

  Accordingly, the present invention has been made to solve these problems, and an object thereof is to provide a communication control device, a communication control system, and a communication control method that can detect and control the occurrence of congestion at an early stage. And

  The communication control device (for example, communication control device 1 described later) according to the first aspect of the present invention is a traffic monitoring unit (for example, traffic monitoring unit 121 described later) that monitors the number of requests transmitted from the terminal to the server. And a congestion detection unit (for example, a congestion detection unit 122 described later) that detects congestion based on the number of requests, and when the congestion detection unit detects the congestion, the server transmits to the terminal And a proxy response unit (for example, a proxy response unit 123 to be described later) that transmits a proxy response in place of the response to be performed.

  According to this invention, since the congestion detection unit detects congestion based on the number of requests, it is possible to detect congestion at an early stage even when a response corresponding to the request is not transmitted from the server. In addition, since the proxy response unit sends a proxy response to the terminal instead of the response sent by the server, the terminal displays an error based on the proxy response and does not receive the response. Retransmission can be suppressed. Therefore, the communication control apparatus can detect and control the occurrence of congestion at an early stage.

  Further, the traffic monitoring unit monitors the number of responses transmitted from the server, and the congestion detection unit detects the congestion based on a ratio between the number of requests and the number of responses. Features.

  According to this invention, since the congestion detection unit detects congestion based on the ratio between the number of requests and the number of responses transmitted from the server, the communication control device responds when a failure or the like occurs in the server. Congestion can be detected when is not transmitted normally. Therefore, the communication control device can detect congestion with higher accuracy than when detecting congestion based only on the number of requests.

  The traffic monitoring unit monitors the number of the terminals that transmitted a request to the server, and the congestion detection unit is based on a ratio between the number of requests and the number of responses, and the number of terminals. The congestion is detected.

  According to the present invention, the communication control device can detect congestion only when the number of terminals transmitting requests to the server is large. Congestion can be detected with higher accuracy without detecting congestion when retransmission is repeated.

  Further, the traffic monitoring unit monitors the number of requests transmitted from some of the plurality of terminals and the number of responses transmitted from the server to the some terminals. It is characterized by.

  Here, there are tens of millions of terminals communicating via the core network of one communication carrier. According to the present invention, the communication control device can detect congestion based on thousands of terminals sampled from tens of millions of terminals, for example, without overloading the communication control device, Congestion can be detected.

The proxy response unit identifies an application managed by the server based on the request, selects a protocol having a congestion control effect for the application from a plurality of protocols, and corresponds to the selected protocol. The proxy response is transmitted.
The proxy response unit identifies a function included in the application based on the request, and selects a protocol having a congestion control effect for the function from the plurality of protocols.
The proxy response unit identifies an operating system of the terminal based on the request, and a protocol having a congestion control effect on a function of the application executed on the operating system among the plurality of protocols. Select.

  According to the present invention, the communication control device transmits a proxy response using a protocol suitable for each function of the application, so that congestion can be controlled effectively. In addition, even if an application performs different communication control for each operating system, the communication control device can effectively control congestion by selecting a protocol having a congestion control effect for each application and for each operating system. Can do.

  Further, the proxy response unit transmits the proxy response based on a ratio between the number of requests monitored by the traffic monitoring unit and the number of responses after transmitting the proxy response instead of the server. The congestion control effect is measured, and a protocol having the congestion control effect is reselected from the plurality of protocols not yet selected based on the congestion control effect.

  According to this invention, the communication control device performs congestion control by selecting a new protocol when there is no congestion control effect by the selected protocol. By doing so, the communication control device performs congestion control using a plurality of protocols even when the congestion control effect by one protocol is small, so that congestion can be controlled more reliably.

  The proxy response unit calculates the number of control target terminals to which the proxy response is to be transmitted, and transmits the proxy response at a time corresponding to a retry time at which the terminal transmits the request at the time of congestion. Based on the number of possible terminals and the number of control target terminals, the number of blocks indicating the set of terminals to which the proxy response is transmitted is calculated, and the number of blocks and the congestion are not generated. A control execution time for executing the control for transmitting the proxy response per block based on a normal retry interval at which the terminal transmits a request, and for each of the blocks over the control execution time The control is executed.

  According to this invention, for example, even if the number of terminals to be controlled is tens of millions, the communication control device can significantly reduce the load on the communication control device by performing processing in a divided manner. . In addition, since the process is divided into a plurality of blocks and the control execution time of each of the plurality of blocks is adjusted based on the normal retry interval, congestion control can be uniformly performed for each of the plurality of blocks. In addition, after controlling each block, the communication control device determines whether or not there is an effect of the control, and controls the next block only when it is determined that there is an effect. Thereby, the influence range at the time of performing control without the effect of suppressing the load can be limited to one block, and execution of the control for other blocks can be prevented.

  A communication control system according to a second aspect of the present invention is a communication control system comprising a terminal and a communication control device that controls communication between the terminal and the server, wherein the terminal requests the server. The communication control device transmits a traffic monitoring unit that monitors the number of requests transmitted from the terminal to the server, a congestion detection unit that detects congestion based on the number of requests, and the congestion detection A communication control device having a proxy response unit that transmits a proxy response instead of a response transmitted by the server to the terminal when the unit detects the congestion.

  A communication control method according to a third aspect of the present invention is a communication control method for controlling communication between a terminal and a server, wherein the number of requests transmitted from the terminal to the server is monitored. Detecting congestion based on the number of requests, and transmitting a proxy response instead of a response transmitted by the server to the terminal when the congestion is detected.

  According to these inventions, the same effects as those of the communication control device according to the first aspect can be obtained.

  According to the present invention, the occurrence of congestion can be detected and controlled at an early stage.

1 is a schematic diagram of a communication control system according to a first embodiment. It is a functional block diagram of the communication control apparatus which concerns on 1st Embodiment. It is a flowchart of the congestion monitoring process which concerns on 1st Embodiment. It is a flowchart of the congestion control process which concerns on 1st Embodiment. It is a flowchart of the control effect measurement process which concerns on 1st Embodiment. It is a figure which shows an example of the effect information according to protocol which concerns on 1st Embodiment. It is a figure which shows the verification result of the control effect measurement process which concerns on 1st Embodiment. It is a figure explaining the change of the traffic at the time of dividing | segmenting into the some block which concerns on 2nd Embodiment, and performing congestion control. It is a flowchart of the congestion control process which concerns on 2nd Embodiment.

Hereinafter, embodiments of the present invention will be described.
<First Embodiment>
[Outline of communication control system S]
FIG. 1 is a schematic diagram of a communication control system S according to the first embodiment.
The communication control system S includes a communication control device 1, a plurality of mobile terminals 2 (mobile terminal 2A, mobile terminal 2B, mobile terminal 2C...), And a plurality of servers 3 (server 3A, server 3B, server 3C). A core network 4, an L1 (layer 1) switch 5, a gateway 6, and the Internet 7.

  The mobile terminal 2 is communicably connected to the server 3 via the core network 4, the L1 switch 5, the gateway 6 and the Internet 7. The mobile terminal 2 is installed with an application that communicates with the server 3. When the application is executed on the mobile terminal 2, the application periodically transmits a request to the server 3. Here, the request indicates all commands transmitted from the mobile terminal 2 to the server 3. If the mobile terminal 2 can receive a response from the server 3, that is, normally, sends a request at a normal retry interval and fails to receive a response that is a response to the request from the server 3, congestion occurs. The request is retransmitted to the server 3 at the retry interval.

  The core network 4 is a basic communication line of a predetermined communication carrier. The L1 switch 5 is a device that duplicates a packet passing through the core network 4 at layer 1 and transmits a request transmitted by the portable terminal 2 to the server 3 and the communication control device 1 and carries a response corresponding to the request. It transmits to the terminal 2 and the communication control apparatus 1.

  The server 3 provides a service related to the application installed in the mobile terminal 2, and is, for example, an application server. When the server 3 receives a request from an application running on the mobile terminal 2, the server 3 transmits a response to the mobile terminal 2.

  The communication control device 1 monitors the traffic in the core network 4 via the L1 switch 5 to detect the occurrence of congestion due to the failure of the server 3 and performs congestion control to suppress the congestion. Specifically, the communication control device 1 monitors traffic in the core network 4, the number of requests transmitted from some portable terminals 2 to the server 3, and the number of requests transmitted from the server 3 to the portable terminal 2. The number of responses is totaled ((1) in FIG. 1). Thereafter, when a failure occurs in any of the servers 3 and the server 3 is down ((2) in FIG. 1), the mobile terminal 2 that has been communicating with the server 3 The request is transmitted ((3) in FIG. 1), and congestion occurs. The communication control device 1 detects congestion caused by the failure of the server 3 based on the total number of requests and the number of responses. When the communication control device 1 detects congestion due to a failure of the server 3, the communication control device 1 transmits a proxy response in place of the server 3 to the request transmitted from the mobile terminal 2 to the server 3 ( (4) of FIG.

[Configuration of Communication Control Device 1]
Then, the function with which the communication control apparatus 1 is provided is demonstrated. FIG. 2 is a functional configuration diagram of the communication control apparatus 1 according to the first embodiment.

The communication control device 1 includes a storage unit 11 and a control unit 12.
The storage unit 11 is configured by, for example, a ROM and a RAM. The storage unit 11 stores various programs for causing the communication control device 1 to function. The storage unit 11 may store a program stored in a storage medium such as an external memory or a program acquired from an external device.

  The control part 12 is comprised by CPU, for example. The control unit 12 comprehensively controls functions related to the communication control device 1 by executing various programs stored in the storage unit 11. The control unit 12 includes a traffic monitoring unit 121, a congestion detection unit 122, and a proxy response unit 123.

  The traffic monitoring unit 121 monitors the number of requests transmitted from the mobile terminal 2 to the server 3, the number of responses transmitted from the server 3, and the number of mobile terminals 2 that transmitted the request to the server 3. The congestion detection unit 122 detects congestion caused by the failure of the server 3 based on the ratio between the number of requests and the number of responses counted by the traffic monitoring unit 121 and the number of request terminals.

  The traffic monitoring unit 121 and the congestion detection unit 122 cooperate to repeatedly execute the congestion monitoring process. Hereinafter, functions of the traffic monitoring unit 121 and the congestion detection unit 122 will be described with reference to flowcharts. FIG. 3 is a flowchart of the congestion monitoring process according to the first embodiment.

  First, the traffic monitoring unit 121 transmits the number of requests transmitted from a part of the plurality of mobile terminals 2 to the server 3 and the server 3 to the part of the mobile terminals. Monitor the number of responses. Specifically, the traffic monitoring unit 121 counts the number of requests transmitted from some portable terminals 2 to the server 3 during a predetermined period (S1). The traffic monitoring unit 121 aggregates the number of responses transmitted from the server 3 to some portable terminals 2 during a predetermined period (S2). Some of these mobile terminals 2 are also referred to as mobile terminals 2 that are monitored. Here, the traffic monitoring unit 121 determines each of the plurality of servers 3 based on the IP address and port number of the server 3 included in the request and the IP address and port number of the server 3 included in the response. , Count the number of requests and the number of responses. That is, the traffic monitoring unit 121 counts the number of requests and the number of responses when the IP address and port number included in the request and response match the IP address and port number of the server 3.

  The congestion detection unit 122 analyzes the response transmitted from the server 3 to the mobile terminal 2 and counts the number of error responses in a predetermined period. The congestion detection unit 122 acquires a threshold value that is stored in advance in the storage unit 11 for detecting the occurrence of a failure in the server 3. The congestion detection unit 122 determines whether or not the number of error responses is larger than a threshold (S3). When the number of error responses is larger than the threshold (when the determination is Yes), the congestion detection unit 122 moves the process to S11, and when the number of error responses is equal to or less than the threshold (when the determination is No), the congestion detection unit 122 performs the process at S4. Move.

  Subsequently, the congestion detection unit 122 calculates the response rate of the server 3 by calculating the ratio between the number of requests and the number of responses collected by the traffic monitoring unit 121 during a predetermined period (S4).

  Subsequently, the congestion detection unit 122 acquires a reference response rate corresponding to the current date and time (S5). Specifically, the storage unit 11 includes each time zone of each day of the week and a reference response rate that is an average response rate of the server 3 when no congestion occurs in each time zone of each day of the week. The associated reference response information is stored. The congestion detection unit 122 refers to the reference response information stored in the storage unit 11 and acquires a reference response rate corresponding to the current time.

  Subsequently, the congestion detection unit 122 determines whether or not the response rate calculated in S4 is lower than the reference response rate acquired in S5 (S6). When the congestion detection unit 122 determines that the response rate is lower than the reference response rate (when the determination is Yes), the congestion detection unit 122 moves the process to S8 and when the response rate is determined to be equal to or higher than the reference response rate (when the determination is No). ), The process is moved to S7.

  If it is determined in S6 that the response rate is equal to or higher than the reference response rate, the congestion detection unit 122 waits for a predetermined time (for example, 1 minute) because no congestion has occurred (S7). After that, the congestion detection unit 122 moves the process to S1. As a result, the traffic monitoring unit 121 aggregates requests and responses, and congestion monitoring is continuously performed.

  In S6, when the congestion detection unit 122 determines that the response rate is lower than the reference response rate, the traffic monitoring unit 121 determines the number of mobile terminals 2 that are sending requests to the server 3, that is, request terminals. Count the numbers.

  Subsequently, the congestion detection unit 122 acquires the number of reference terminals corresponding to the current date and time (S9). Specifically, the storage unit 11 associates each time zone of each day of the week with a reference terminal number indicating an average number of request terminals when congestion does not occur in each time zone of each day of the week. Reference terminal number information is stored. The congestion detection unit 122 refers to the reference terminal number information stored in the storage unit 11 and acquires the reference terminal number corresponding to the current time.

  Subsequently, the congestion detection unit 122 determines whether or not the number of request terminals counted in S8 is larger than the reference terminal number acquired in S9 (S10). If the congestion detection unit 122 determines that the number of request terminals is greater than the reference terminal number (when the determination is Yes), the congestion detection unit 122 detects that congestion has occurred and moves the process to S11. If the congestion detection unit 122 determines that the number of request terminals is equal to or less than the number of reference terminals (when the determination is No), the congestion detection unit 122 moves the process to S7.

  Subsequently, the proxy response unit 123 executes a congestion control process (S11). This process will be described later.

  Next, the function of the proxy response unit 123 will be described. When the congestion detection unit 122 detects congestion, the proxy response unit 123 executes a congestion control process, thereby transmitting a proxy response instead of the response transmitted by the server 3 to the mobile terminal 2. Hereinafter, the details of the congestion control process executed by the proxy response unit 123 will be described with reference to flowcharts. FIG. 4 is a flowchart of the congestion control process according to the first embodiment.

  First, the proxy response unit 123 specifies a congestion service that is a congested service based on the destination included in the request transmitted from the mobile terminal 2 to the server 3, that is, the IP address and port number of the server 3. (S21). Here, the service indicates a service related to an application managed in the server 3.

  Specifically, the storage unit 11 stores a management table in which the IP address and port number of the server 3 are associated with the names of applications managed by the server 3. The proxy response unit 123 refers to the management table, identifies the application name corresponding to the IP address and port number included in the request, and identifies the application managed by the server 3 based on the request.

  Here, in the management table, the IP address, the port number, and the function included in the application may be associated with each other, and the proxy response unit 123 may specify the function included in the application based on the request. In the management table, the IP address, the port number, and the type of the operating system (hereinafter referred to as OS) of the mobile terminal 2 are associated with each other, and the proxy response unit 123 determines the OS of the mobile terminal 2 based on the request. May be specified. Here, the functions included in the application are, for example, functions related to a call function, a message transmission function, and background processing, and are functions that transmit a request to the server 3.

  Subsequently, the proxy response unit 123 acquires a retry interval at the time of congestion (S22). Specifically, in the storage unit 11, the application name managed by the server 3 is associated with the retry interval (retry interval at the time of congestion) of a request transmitted from the portable terminal 2 to the server 3 at the time of congestion. Retry information is stored. The proxy response unit 123 acquires the retry interval at the time of congestion of the application specified at S21 based on the retry information at the time of congestion.

  Subsequently, the proxy response unit 123 determines the congestion control execution time based on the retry interval at the time of congestion (S23). Specifically, the proxy response unit 123 sets the retry interval at the time of congestion as the congestion control execution time. As described above, when the retry interval at the time of congestion is the congestion control execution time, the communication control device 1 can receive a request from the mobile terminal 2 one or more times during the execution of the congestion control. Then, when the communication control device 1 transmits a proxy response to the request, the mobile terminal 2 displays an error or the like, and the transmission of the request to the server 3 is suppressed for a certain time.

  Subsequently, the proxy response unit 123 specifies the mobile terminal 2 that is transmitting the request to the server 3 that manages the application (congestion service) specified in S21 as the control target terminal that is the target of transmitting the proxy response. The number of control target terminals, which is the number of control target terminals, is acquired (S24). Specifically, the L1 switch 5 duplicates a request transmitted from the mobile terminal 2 to the server 3 via the core network 4 and supplies the request to the server 3 and the communication control device 1. Then, the proxy response unit 123 identifies the number of control target terminals by counting the number of requests.

Subsequently, the proxy response unit 123 selects the control with the highest priority among the unexecuted controls (S25). Selection of control with high priority is performed as follows.
First, the storage unit 11 stores protocol-specific effect information in which an application name and information indicating a congestion control effect related to each application of a plurality of protocols are associated with each other. FIG. 6 is a diagram illustrating an example of the effect information by protocol according to the first embodiment. FIG. 6 shows, as an example, information indicating the congestion control effect of each protocol for the applications A to F when the OS type of the mobile terminal 2 is “OS1”.

  For example, “ICMP1 (TYPE3)” indicates “destination unreachable” in the ICMP protocol, and “ICMP2 (TYPE4)” indicates “start point suppression” in the ICMP protocol. “SYN1 (FIN)” indicates a “FIN (end) flag” in the TCP protocol, and “SYN2 (RST)” indicates an “RST (reset) flag” in the TCP protocol. FIG. 6 shows information indicating the congestion control effect of each protocol for the applications A and B when the OS type of the mobile terminal 2 is “OS2”. Here, four types of information “情報”, “◯”, “Δ”, and “×” are shown as information indicating the congestion control effect of each protocol. For example, “◎” indicates that the congestion control effect is the highest, and “×” indicates that the congestion control effect is the lowest. Information indicating the congestion control effect of each protocol is used for selecting a protocol by the proxy response unit 123. As shown in FIG. 6, even if the application A and the application B are the same application, protocols having a high congestion control effect are different for each type of OS of the mobile terminal 2.

  The proxy response unit 123 refers to the protocol-specific effect information and selects a protocol having a congestion control effect for the application from among a plurality of protocols. Specifically, the proxy response unit 123 selects a protocol having the highest congestion control effect for the application among a plurality of protocols. Here, the proxy response unit 123 may select a protocol having a congestion control effect for each function of the application. Further, the proxy response unit 123 may select a protocol that has a congestion control effect on the function of the application executed on the OS.

  Subsequently, the proxy response unit 123 determines whether to wait for a response from the server 3 (S26). Specifically, for each server 3, a setting value indicating whether to wait for a response is stored in advance in the storage unit 11, and the proxy response unit 123 waits for a response from the server 3 based on the setting value. Determine whether or not. The proxy response unit 123 moves the process to S27 when waiting for a response from the server 3, and moves the process to S28 when not waiting for a response from the server 3.

  Subsequently, the proxy response unit 123 checks whether there is a control target terminal that has received a response from the server 3 (S27). If the response of the response can be confirmed, the proxy response unit 123 determines that the congestion can be resolved by the server 3 recovering and transmitting the response normally, and the process of this flowchart ends. If the response cannot be confirmed, the proxy response unit 123 moves the process to S28.

  Subsequently, the proxy response unit 123 executes congestion control over the congestion control execution time determined in S23 (S28). Specifically, when the proxy response unit 123 detects that the traffic monitoring unit 121 has transmitted a request to the server 3 from the mobile terminal 2 that is the control target terminal within the congestion control execution time determined in S23, A proxy response is transmitted to the portable terminal 2. In this control, for the mobile terminal 2 to be monitored, the execution of congestion control may be delayed in order to accurately perform the control effect measurement described later.

The proxy response unit 123 performs a control effect measurement process and measures the effect of the control executed in S27 (S29). Details of the control effect measurement process will be described later.
Subsequently, the proxy response unit 123 determines whether or not the control effect measured in S29 is higher than a threshold value (S30). If the proxy response unit 123 determines that the effect ratio is higher than the threshold value (when the determination is Yes), the proxy response unit 123 ends the congestion control process. In addition, when the proxy response unit 123 determines that the effect ratio is equal to or less than the threshold (when determination is No), the proxy response unit 123 proceeds to S25. By moving the process to S25, the proxy response unit 123 reselects a protocol having a congestion control effect from a plurality of protocols that have not yet been selected by referring to the protocol-specific effect information.

  Next, details of the control effect measurement process will be described. In the control effect measurement process, the proxy response unit 123 transmits a response instead of the server 3 and then transmits a proxy response based on the ratio between the number of requests monitored by the traffic monitoring unit 121 and the number of responses. Measure the congestion control effect by. FIG. 5 is a flowchart of the control effect measurement process according to the first embodiment.

  First, the proxy response unit 123 acquires a congestion service response rate in the monitored portable terminal 2 (S31). Specifically, the proxy response unit 123 includes the number of requests transmitted from some mobile terminals 2 that are monitored by the traffic monitoring unit 121 and the number of responses transmitted to some mobile terminals 2. Get response rate based on. Hereinafter, the response rate acquired in S31 is referred to as response rate (a).

  Subsequently, the proxy response unit 123 acquires the response rate of the congestion service in the terminal after control (S32). Specifically, the proxy response unit 123 calculates the response rate based on the number of requests transmitted from the control target terminal specified in S24 of the congestion control process and the number of responses transmitted to the control target terminal. get. Hereinafter, the response rate acquired in S32 is referred to as response rate (b).

  Subsequently, the proxy response unit 123 determines whether or not the response rate (a) acquired in S31 is smaller than the response rate (b) acquired in S32 (S33). When it is determined that the response rate (a) is smaller than the response rate (b) (when the determination is Yes), the proxy response unit 123 determines “effective” (S34). When it is determined that the response rate (a) is equal to or higher than the response rate (b) (when the determination is No), it is determined that there is no effect (S38).

If the proxy response unit 123 determines that “there is an effect”, the proxy response unit 123 calculates an effect ratio based on the response rate (a) and the response rate (b) (S35). Specifically, the proxy response unit 123 calculates a ratio of the response rate (b) to the response rate (a), that is, an increase rate of the response rate as an effect rate.
Subsequently, the proxy response unit 123 records the control method and the effect ratio (S36). Here, the control method indicates the type of control based on each of the plurality of protocols, and the proxy response unit 123 stores the numerical value of the effect ratio in the storage unit 11 in association with the plurality of protocols.

  Subsequently, the proxy response unit 123 changes the priority of the control method based on the calculated effect ratio (S37). Specifically, the proxy response unit 123 has four types of congestion control effects of “◎”, “◯”, “Δ”, and “×” based on the numerical value of the effect ratio recorded in the storage unit 11. The effect information for each protocol is updated. As a result, the protocol-specific effect information can be dynamically changed, and the protocol priority can be dynamically changed in the next congestion control process.

  Subsequently, the proxy response unit 123 calculates a control effect based on the response rate (a) and the response rate (b) (S39). Specifically, the proxy response unit 123 calculates a numerical value indicating the control effect based on the information indicating the congestion control effect classified based on the numerical value of the effect ratio. For example, when the information indicating the congestion control effect is “◎”, the control effect indicates the maximum value, and when it is determined “no effect” in S38, the control effect indicates the minimum value.

[inspection result]
Next, verification results for the above-described congestion control processing will be shown.
FIG. 7 is a diagram illustrating a verification result of the control effect measurement process according to the first embodiment. In the verification, in the state where one mobile terminal 2 is connected to the server 3 via the L2 switch and the router, the communication between the mobile terminal 2 and the server 3 is cut off by the router, and the server 3 is down in a pseudo manner. Simulated the condition. And the communication control apparatus 1 performed the proxy response to the portable terminal 2 via L2 switch.

  For example, in the background process of application A, when the proxy response is not transmitted, the request is generated 8 times per minute, but the communication control device 1 sets the SYN / ACK flag of the TCP protocol as the proxy response. Sent, and then the FIN flag was sent. As a result, 1 minute requests were reduced to 3 times.

  In addition, in the application B application startup process, when a proxy response is not transmitted, eight requests are generated per minute. However, the communication control apparatus 1 uses ICMP protocol TYPE 3 (destination arrival) as a proxy response. Impossible). As a result, 1 minute requests were reduced to 2 times.

  Further, in the background process of application C, when a proxy response is not transmitted, the request is generated 12 times per minute, but the communication control device 1 transmits a DNS error (inquiry unknown) as a proxy response. did. As a result, requests for 1 minute were reduced to one time.

[Effect in the first embodiment]
As described above, the communication control device 1 according to the first embodiment includes the traffic monitoring unit 121 that monitors the number of requests transmitted from the mobile terminal 2 to the server 3 and the congestion that detects congestion based on the number of requests. When the congestion detection part 122 detects congestion, the detection part 122 and the proxy response part 123 which transmits the proxy response instead of the response which the server 3 transmits with respect to the portable terminal 2 are provided.

  Since the congestion detection unit 122 detects congestion based on the number of requests, the communication control device 1 can detect congestion at an early stage even when a response corresponding to the request is not transmitted from the server 3. Further, since the proxy response unit 123 transmits a proxy response instead of the response transmitted by the server 3 to the mobile terminal 2, the mobile terminal 2 displays an error based on the proxy response and the server 3 It is possible to suppress retransmission of requests due to not receiving a response from. Therefore, the communication control apparatus 1 can detect and control the occurrence of congestion at an early stage. The control here refers to, for example, thinning out the number of requests by 30% or stopping the request itself.

<Second Embodiment>
[Perform congestion control for the control target terminal separately]
Next, the second embodiment will be described. The second embodiment differs from the first embodiment in that the proxy response unit 123 divides and executes the congestion control for the control target terminal in the congestion control process, and is the same in other points.

  The proxy response unit 123 according to the second embodiment calculates the number of control target terminals (number of control target terminals) that are the mobile terminals 2 for congestion control. Here, the number of control target terminals is, for example, tens of millions, and the communication control apparatus 1 substitutes for all the control target terminals in a time corresponding to a retry interval (for example, 10 seconds) at the time of congestion. The response cannot be sent.

  Here, when the mobile terminal 2 can receive a response from the server 3, that is, during normal times, it transmits a request at a normal retry interval, and when it cannot receive a response from the server 3, it tries to retry during congestion. The request is retransmitted to the server 3 at intervals. Then, the mobile terminal 2 waits for a time corresponding to a normal retry interval (for example, 1 minute) by receiving the proxy response, and then resends the request to the server 3.

  That is, when the mobile terminal 2 receives a proxy response at the time of detecting congestion, it waits for a time corresponding to the normal retry interval, so the retry interval changes from the retry interval at the time of congestion to the normal retry interval. Therefore, if a proxy response is transmitted to all control target terminals in a time corresponding to the normal retry interval, the retry interval of all control target terminals is changed from the retry interval at the time of congestion to the normal retry interval. become.

  Therefore, the proxy response unit 123 determines the normal retry interval as the congestion control execution time for performing the congestion control. Then, the proxy response unit 123 determines the control target terminal based on the number of terminals that can transmit a proxy response in the time corresponding to the retry interval at the time of congestion when the mobile terminal 2 transmits a request at the time of congestion, and the number of control target terminals. Among them, a plurality of blocks including some control target terminals are generated. Here, the block is a set of some control target terminals that are targets of transmitting a proxy response in a time corresponding to a retry interval at the time of one congestion. For example, when the number of control target terminals is 6 million and the number of terminals that can be processed at one time (the number of terminals that can transmit a proxy response in a time corresponding to the retry interval at the time of congestion) is 1 million, the number of blocks Is determined to be 6. The proxy response unit 123 transmits a proxy response in order for each of the plurality of blocks.

  FIG. 8 is a diagram for explaining a change in traffic when congestion control is performed by dividing into a plurality of blocks according to the second embodiment. For example, as shown in FIG. 8, it is assumed that the control target terminal is divided into six blocks from block A to block F. Further, it is assumed that the retry interval at the time of congestion is 10 seconds and the retry interval at the normal time is 60 seconds.

  The proxy response unit 123 determines the control execution time per block based on the congestion control execution time and the number of blocks. Specifically, the proxy response unit 123 determines the control execution time per block by dividing the congestion control execution time by the number of blocks using the normal retry interval as the congestion control execution time.

  In the example shown in FIG. 8, the control execution time per block is determined to be 10 seconds. When the control execution time per block is longer than the time corresponding to the retry interval during congestion, the proxy response unit 123 sets the control execution time per block as the time corresponding to the retry interval during congestion. Also good. For example, when the time corresponding to the retry interval at the time of congestion is 10 seconds and the control execution time per block is 15 seconds, the proxy response unit 123 is included in one block by performing control over 10 seconds. The proxy response can be transmitted to all the control target terminals within a time corresponding to the retry interval at the time of congestion.

  Subsequently, the proxy response unit 123 transmits a proxy response for the determined control execution time to the control target terminal included in each block. In the example illustrated in FIG. 8, the proxy response unit 123 performs congestion control on the control target terminals included in the block A during 0 to 10 seconds. Then, the traffic corresponding to the control target terminal included in the block A is suppressed for a time (60 seconds) corresponding to the retry interval at the normal time. Subsequently, when the proxy response unit 123 performs the congestion control for the control target terminal included in the block B within 10 seconds to 20 seconds, the traffic corresponding to the control target terminal included in the block B is the normal time. It is suppressed over a time corresponding to the retry interval. Similarly, the proxy response unit 123 performs congestion control every 10 seconds in the order of block C, block D, block E, and block F during 20 to 60 seconds. As a result, the proxy response unit 123 can suppress traffic in the block processing order.

  Subsequently, the congestion control process of the proxy response unit 123 according to the second embodiment will be described with reference to a flowchart. FIG. 9 is a flowchart of the congestion control process according to the second embodiment.

First, the proxy response unit 123 specifies a congestion service (application) that is a service during congestion based on the destination included in the request transmitted from the mobile terminal 2 to the server 3 (S41).
Subsequently, the proxy response unit 123 acquires the retry interval at the normal time (S42). More specifically, the storage unit 11 stores normal retry information in which the application name managed by the server 3 is associated with the normal retry interval. The proxy response unit 123 acquires a normal retry interval based on the normal retry information.

Subsequently, the proxy response unit 123 determines the congestion control execution time based on the normal retry interval acquired in S42 (S43). Specifically, the proxy response unit 123 sets the normal retry interval as the congestion control execution time.
Subsequently, the proxy response unit 123 acquires the number of terminals to be controlled and the number of terminals that can be processed at one time (S44). For example, the proxy response unit 123 acquires the number of portable terminals 2 that can transmit a proxy response within a time corresponding to a retry interval at the time of congestion as the number of control target terminals.

Subsequently, the proxy response unit 123 determines the number of blocks based on the number of terminals to be controlled and the number of terminals that can be processed at a time acquired in S44.
Subsequently, the proxy response unit 123 calculates a control execution time per block based on the congestion control execution time determined in S43 and the number of blocks determined in S45 (S46).

Subsequently, the proxy response unit 123 selects the control with the highest priority among the unexecuted controls (S47).
Subsequently, the proxy response unit 123 determines whether to wait for a response from the server 3 (S48). The proxy response unit 123 moves the process to S49 when waiting for a response from the server 3, and moves the process to S50 when not waiting for a response from the server 3.
Subsequently, the proxy response unit 123 checks whether there is a control target terminal that has received a response from the server 3 (S49).

  Subsequently, the proxy response unit 123 executes congestion control for each block (S50). If the control execution time calculated in S46 is longer than the time corresponding to the retry interval at the time of congestion, the proxy response unit 123 transmits the proxy response over the time corresponding to the retry interval at the time of congestion. If the control execution time per block calculated in S46 is shorter than the time corresponding to the retry interval at the time of congestion, the proxy response unit 123 transmits a proxy response over the control execution time calculated in S46. When the control execution time per block is shorter than the time corresponding to the retry interval at the time of congestion, a proxy response is transmitted to a part of the control target terminals corresponding to one block.

  More specifically, the proxy response unit 123 transmits a request to the server 3 from the portable terminal 2 that is a control target terminal corresponding to the block that is being controlled by the traffic monitoring unit 121 within the control execution time calculated in S46. When it is detected that this has been done, a proxy response is transmitted to the portable terminal 2. Here, if the control execution time per block is shorter than the retry interval at the time of congestion, proxy responses cannot be sent to all control target terminals included in one block, but proxy is performed at a certain rate. Since the response can be transmitted, the traffic related to some of the control target terminals included in the block 1 can be suppressed.

  The proxy response unit 123 performs a control effect measurement process and measures the effect of the control executed in S50 (S51). Since the control effect measurement process is the same as the control effect measurement process of the first embodiment, the description thereof is omitted.

  Subsequently, the proxy response unit 123 determines whether or not the control effect measured in S51 is higher than a threshold value (S52). If the proxy response unit 123 determines that the control effect is higher than the threshold (if the determination is Yes), the proxy response unit 123 proceeds to S53. In addition, when the proxy response unit 123 determines that the control effect is equal to or less than the threshold value (when the determination is No), the proxy response unit 123 proceeds to S47.

  Subsequently, the proxy response unit 123 determines whether or not control has been performed on all processing target terminals, that is, whether or not congestion control processing has been performed on all blocks (S53). If the proxy response unit 123 determines that control has been performed on all the processing target terminals (if the determination is Yes), the proxy response unit 123 ends the congestion control process and does not control all the processing target terminals. When it determines (when determination is No), a process is moved to S48.

[Effects of Second Embodiment]
As described above, the proxy response unit 123 according to the second embodiment calculates the number of control target terminals that are targets for transmitting a proxy response, and corresponds to a retry time during congestion when the mobile terminal 2 transmits a request during congestion. Based on the number of terminals that can send a proxy response in time and the number of terminals to be controlled, the number of blocks indicating the set of mobile terminals 2 to which proxy responses are sent is calculated, and the number of blocks and congestion occur If not, the control execution time for executing the control per block is specified based on the normal retry interval at which the mobile terminal 2 transmits the request, and the control is executed for each block over the control execution time. To do.

  According to this invention, for example, even if the number of terminals to be controlled is tens of millions, the communication control device 1 significantly reduces the load on the communication control device 1 by performing processing in a divided manner. Can do. In addition, since the process is divided into a plurality of blocks and the control execution time of each of the plurality of blocks is adjusted based on the normal retry interval, congestion control can be uniformly performed for each of the plurality of blocks.

<Third Embodiment>
[Determination of congestion based on the number of requests]
Subsequently, a third embodiment will be described. The third embodiment is different from the first embodiment in that the congestion detection unit 122 determines congestion based on the number of requests without using the number of responses, and is the same in other points.

  The traffic monitoring unit 121 according to the third embodiment monitors the number of requests transmitted from the mobile terminal 2 to the server 3 during the predetermined period and the number of mobile terminals 2 that transmitted the request to the server 3. Specifically, the traffic monitoring unit 121 counts the number of requests transmitted from the mobile terminal 2 to the server 3 during a predetermined period based on the IP address and port number of the server 3 included in the request as destination information. To do. In addition, the traffic monitoring unit 121 counts the number of mobile terminals 2 that transmitted the request to the server 3 during a predetermined period based on the IP address and port number of the mobile terminal 2 included as transmission source information in the request.

  The congestion detection unit 122 detects congestion caused by the failure of the server 3 based on the number of requests counted by the traffic monitoring unit 121 and the number of request terminals. Specifically, the congestion detection unit 122 calculates the rate of change in the number of requests counted by the traffic monitoring unit 121 over a plurality of predetermined periods. Then, the congestion detection unit 122 detects congestion when the rate of change exceeds a threshold value, that is, when the number of requests rapidly increases.

[Effect in the third embodiment]
As described above, the communication control device 1 according to the third embodiment includes the traffic monitoring unit 121 that monitors the number of requests transmitted from the mobile terminal 2 to the server 3 and the congestion that detects congestion based on the number of requests. When the congestion detection part 122 detects congestion, the detection part 122 and the proxy response part 123 which transmits the proxy response instead of the response which the server 3 transmits with respect to the portable terminal 2 are provided.

  According to the present invention, it is not necessary for the traffic monitoring unit 121 to tabulate responses, so that the load on the communication control device 1 can be reduced as compared with the case of summing up responses. In addition, since the congestion detection unit 122 detects congestion based only on the number of requests, it is possible to detect congestion early even if a response corresponding to the request is not transmitted from the server.

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment. It will be apparent to those skilled in the art that various modifications or improvements can be added to the above embodiment. In particular, the specific embodiments of the distribution / integration of the devices are not limited to those illustrated above, and all or a part thereof may be added in arbitrary units according to various additions or according to functional loads. It can be configured functionally or physically distributed and integrated.

DESCRIPTION OF SYMBOLS 1 ... Communication control apparatus, 11 ... Memory | storage part, 12 ... Control part, 121 ... Traffic monitoring part, 122 ... Congestion detection part, 123 ... Proxy response part, 2 ... Mobile terminal, 3 ... server, 4 ... core network, 5 ... L1 switch, 6 ... gateway, 7 ... Internet, S ... communication control system

Claims (12)

A traffic monitoring unit that monitors the number of requests sent from the terminal to the server;
A congestion detector that detects congestion based on the number of requests;
A proxy response unit that transmits a proxy response in place of a response transmitted by the server to the terminal when the congestion detection unit detects the congestion;
Communication control device. The proxy response unit, when the congestion detection unit detects the congestion, transmits the proxy response having a congestion control effect to the application corresponding to the request to the terminal.
The communication control apparatus according to claim 1. The traffic monitoring unit monitors the number of responses transmitted from the server,
The congestion detection unit detects the congestion based on a ratio between the number of requests and the number of responses;
The communication control apparatus according to claim 1 or 2 . The traffic monitoring unit monitors the number of the terminals that transmitted a request to the server,
The congestion detection unit detects the congestion based on a ratio between the number of requests and the number of responses, and the number of terminals.
The communication control apparatus according to claim 3 . The traffic monitoring unit monitors the number of requests transmitted from some of the plurality of terminals and the number of responses transmitted from the server to the some terminals;
The communication control apparatus according to claim 3 or 4 . The proxy response unit identifies an application managed by the server based on the request, selects a protocol having a congestion control effect for the application from a plurality of protocols, and the proxy corresponding to the selected protocol Send a response,
The communication control apparatus according to any one of claims 3 to 5 . The proxy response unit identifies a function included in the application based on the request, and selects a protocol having a congestion control effect for the function from the plurality of protocols.
The communication control apparatus according to claim 6 . The proxy response unit identifies an operating system of the terminal based on the request, and selects a protocol having a congestion control effect for the function of the application executed on the operating system from the plurality of protocols. To
The communication control apparatus according to claim 7 . The proxy response unit transmits the proxy response instead of the server, and then transmits the proxy response based on a ratio between the number of requests monitored by the traffic monitoring unit and the number of responses. Measuring a congestion control effect, and reselecting a protocol having the congestion control effect from the plurality of protocols not yet selected based on the congestion control effect;
The communication control apparatus according to any one of claims 6 to 8 . The proxy response unit
Calculate the number of control target terminals to which the proxy response is transmitted,
Based on the number of terminals that can transmit the proxy response in a time corresponding to a retry time during congestion when the terminal transmits the request at the time of congestion, and the number of terminals to be controlled, the target of transmitting the proxy response Calculate the number of blocks representing a set of terminals,
Specify the control execution time for executing the control for transmitting the proxy response per block based on the number of blocks and the normal retry interval at which the terminal transmits a request when the congestion does not occur And
For each of the blocks, execute the control over the control execution time.
The communication control apparatus according to any one of claims 1 to 9 . A communication control system comprising a terminal and a communication control device that controls communication between the terminal and the server,
The terminal sends a request to the server;
The communication control device includes:
A traffic monitoring unit that monitors the number of requests transmitted from the terminal to the server;
A congestion detector that detects congestion based on the number of requests;
A communication control device having a proxy response unit that transmits a proxy response instead of a response transmitted by the server to the terminal when the congestion detection unit detects the congestion;
Comprising
Communication control system. A communication control method for controlling communication between a terminal and a server,
Monitoring the number of requests sent from the terminal to the server;
Detecting congestion based on the number of requests;
A proxy response instead of a response sent by the server to the terminal when the congestion is detected, and
Communication control method.

Images