JP2006319747A - Congestion control method, congestion control apparatus, tagging apparatus, and disposal apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technology for constantly realizing congestion control without losing the convenience of users legitimately using a congestion control system. <P>SOLUTION: A congestion control apparatus 1 includes a packet analysis section 20, a bad flow prescribing section 22, and a communication control section 26, the packet analysis section 20 executes a procedure of extracting a flow from a tagged packet in an overflow queue of a preferential disposal apparatus 16 according to flow identification information, the bad flow particularizing section 22 executes a procedure for prescribing whether or not the extracted flow is a bad flow by collating characteristic information of the extracted flow with characteristic information of bad flows stored in a storage means, and the communication control section 26 executes a procedure for applying communication control to the prescribed bad flow. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a congestion control method, a congestion control device, a tagging device, and a discarding device.

  As a conventional communication control method such as bandwidth control for suppressing congestion, which is a congestion state on the network, there is a priority discard method (see Non-Patent Document 1). The priority discard method performs communication control by discarding a packet with a discard priority tag. The tag is attached to a packet of a flow whose flow rate or burst size exceeds a threshold value.

The Internet Engineering Task Force (IETF), “An Architecture for Differentiated Services”, [online], [searched on April 14, 2005], Internet <URL: http://www.ietf.org/rfc/rfc2475. txt>

  However, in the conventional communication control method based on tagging, the same communication control processing is performed not only for the malignant flow that causes congestion but also for all flows. The disposal has been performed, and the convenience of the user who is using it properly has been significantly impaired.

  The malignant flow impairs the fairness of network usage among users. The malignant flow is often generated by an application that performs communication using a bandwidth that is free on a daily basis. And even if a business operator performs equipment augmentation, the band is occupied again by some users. Therefore, the difference between the usage band of the user who generates the malignant flow and the usage band of the other user is increased.

  Also, even if the administrator manually controls the malignant flow temporarily, the application that is the source of the malignant flow continues to generate the malignant flow. It is necessary to constantly monitor the malignant flow.

  Therefore, the main object of the present invention is to solve the above-described problems and to constantly realize congestion control without impairing the convenience of a user who is properly using it.

  In order to solve the above problems, the present invention includes a tagging device that tags a packet of a flow to be relayed, a discarding device that discards the packet in an overflow queue with reference to the tag, and a communication device that relays the flow. A congestion control method for controlling congestion occurring in a network constituted by a congestion control device, wherein the congestion control device includes a packet analysis unit, a malicious flow identification unit, and a communication control unit, and the packet analysis unit A procedure for extracting a flow from a packet in the overflow queue according to flow identification information, and the malignant flow specifying unit includes the extracted flow feature information and the malignant flow feature information stored in a storage unit. The procedure for collating and identifying whether the extracted flow is a malignant flow and the communication control unit are identified. And to execute a procedure of communication control malignant flow, a.

  Thereby, the control of congestion can be realized constantly without impairing the convenience of the user who is normally using.

  The present invention is characterized in that the congestion control device has a control order specifying unit, and the control order specifying unit executes a procedure for defining the order of malignant flows controlled by the communication control unit.

  Thereby, it is possible to suppress excessive control of a flow that has little influence on congestion.

  The present invention is characterized in that the tagging device performs tagging so as to preferentially discard a flow having a long flow duration.

  As a result, communication control can be performed effectively when quality degradation over a long period of time occurs.

  The present invention is characterized in that the tagging device performs tagging by changing a value according to a communication amount of a flow from a communication start time.

  As a result, it is possible to identify a flow having a high rate instantaneously and a flow having a constant high rate for a predetermined period of time for tagging, and to discard a packet with a priority.

  The present invention is characterized in that the tagging device measures the traffic of the flow for each packet that is a specific transmission destination or a specific transmission source.

  Thereby, it is also possible to identify a flow whose connection destination changes autonomously and distributedly. For example, when a bandwidth is available, a flow generated by an application such as P2P that performs communication using the bandwidth correspondingly tends to always have a high rate. These flows have autonomous decentralization in which the connection destination is frequently changed during a predetermined time to establish a session with an unspecified number of other parties.

  The present invention is characterized in that the discard device discards a packet in an overflow queue with reference to characteristic information of a flow to which the tag and the packet belong.

  As a result, when there is a difference between the design bandwidth and the actual usage status, it is possible to keep the user's throughput as high as possible by discarding only the minimum necessary packets when performing communication control.

  The present invention is characterized in that the malicious flow identification unit counts the number of packets based on the header information of the packets stored in the overflow queue, and acquires the extracted flow characteristic information.

  Thereby, since all the packets are counted, a malignant flow can be specified accurately.

  The present invention is characterized in that the malicious flow identification unit counts the number of packets based on header information and tag values of packets stored in the overflow queue, and acquires extracted flow feature information. To do.

  Thereby, referring to only the packets that are likely to be malignant flows with reference to the tag value, the number of packets to be counted can be reduced, and the malignant flows can be quickly identified.

  The present invention is characterized in that the communication control unit repeats confirmation of congestion status and communication control until congestion is alleviated.

  Thus, by referring to the latest congestion state in the process of performing communication control of flows in order, it is possible to identify a malignant flow and to avoid communication control of a flow that is not a malignant flow.

  The present invention is characterized in that, after the communication control unit performs communication control of a plurality of flows by a predetermined amount, the communication control is performed for each flow according to the order.

  Thereby, congestion can be quickly alleviated by narrowing a plurality of flows by a predetermined amount.

  The present invention is characterized in that the communication control unit performs communication control in a stepwise manner according to a congestion state.

  Accordingly, since the flow rate is determined while checking the congestion state, it is possible to prevent communication control more than necessary, and to keep the user's throughput as high as possible.

  The present invention is characterized in that the communication control unit acquires a congestion state by a monitoring packet.

  As a result, the flow rate suitable for mitigating congestion is known, and as a result, only the minimum communication control necessary for mitigating congestion is performed to avoid excessively narrowing the bandwidth of the malicious flow and maximizing the user throughput. It can be kept as high as possible.

  The congestion control method according to any one of claims 2 to 8, wherein the communication control unit acquires a congestion state based on a packet loss rate of the communication device.

  As a result, the flow rate suitable for mitigating congestion is known, and as a result, only the minimum communication control necessary for mitigating congestion is performed to avoid excessively narrowing the bandwidth of the malicious flow and maximizing the user throughput. It can be kept as high as possible.

  The present invention occurs in a network including a tagging device that tags a packet of a flow to be relayed, a discarding device that discards the packet in an overflow queue with reference to the tag, and a communication device that relays the flow A congestion control device for controlling congestion, a packet analysis unit for extracting a flow from a packet in the overflow queue according to flow identification information, characteristic information of the extracted flow, and a malicious flow stored in a storage unit A malignant flow identification unit that identifies whether the extracted flow is a malignant flow, and a communication control unit that performs communication control of the identified malignant flow. To do.

  Thereby, the control of congestion can be realized constantly without impairing the convenience of the user who is normally using.

  The present invention relates to a tagging device that tags a packet, a discard device that refers to the tag and discards the packet in an overflow queue, a communication device that relays a flow, and a flow from the packet in the overflow queue according to flow identification information And the feature information of the extracted flow and the feature information of the malignant flow stored in the storage means are collated to determine whether the extracted flow is a malignant flow, and the specified The tagging device used in a congestion control system including a congestion control device for communication control of a malignant flow, characterized in that it tags a packet of a flow to be relayed.

  Thereby, the control of congestion can be realized constantly without impairing the convenience of the user who is normally using.

  The present invention extracts a flow from a tagging device for tagging a packet of a flow to be relayed, a discarding device having an overflow queue, a communication device for relaying a flow, and a packet of the overflow queue according to flow identification information, The extracted flow feature information and the malignant flow feature information stored in the storage means are collated to identify whether or not the extracted flow is a malignant flow, and communication control of the identified malignant flow is performed. A discard device used in a congestion control system including a congestion control device that discards a packet in an overflow queue with reference to a packet tag.

  Thereby, the control of congestion can be realized constantly without impairing the convenience of the user who is normally using.

  Thereby, the control of congestion can be realized constantly without impairing the convenience of the user who is normally using.

  Hereinafter, an embodiment of a congestion control system to which the present invention is applied will be described in detail with reference to the drawings. First, the configuration of the congestion control system of this embodiment will be described with reference to FIGS.

  FIG. 1 is a configuration diagram showing a congestion control system. The control target network 10 includes a communication device 12 that relays a flow, a tagging device 14 that tags a packet of the flow, a priority discard device 16 (discard device in claims) that discards the packet to an overflow queue based on the tag, And the user terminal 18 which generates a flow is connected.

  The tagging device 14 and the priority discarding device 16 are preferably devices that also operate as routers (communication devices 12), each having a transfer function for realizing the flow shown in FIG. The tagging device 14 is installed upstream of the priority discard device 16 that discards packets. For this reason, the tagging device 14 is preferably an edge router located at the boundary with another network.

  Note that the congestion control device 1 that controls congestion that has occurred in the control target network 10 is configured as a computer that includes at least a memory serving as storage means used when performing arithmetic processing, and an arithmetic processing device that performs the arithmetic processing. The The memory is constituted by a RAM (Random Access Memory) or the like. Arithmetic processing is realized by an arithmetic processing unit configured by a CPU (Central Processing Unit) executing a program on a memory.

  FIG. 2 is an explanatory diagram showing tagging processing by the tagging device 14. The tagging device 14 tags the flow that passes. Here, the tagging device 14 attaches a tag so as to conform to the tagging rule based on the flow measurement result. For example, tagging that conforms to the rule of “tagging excess from 10 Mbps per flow” tags an excess 40 Mbps packet for a 50 Mbps flow.

  FIG. 3 is a block diagram showing a tagged packet. Note that the number of tags attached to a packet may be one, or a plurality of tag sets in which the order is defined. Here, the tag order determines a priority to be referred to when determining whether or not to discard a packet. For example, the first tag (priority discard tag) is referred to more preferentially than the second tag (additional information tag). The number of tags constituting the tag set is not limited to two and may be an arbitrary number.

  FIG. 4 is an explanatory diagram showing the priority discarding device 16. The priority discard device 16 monitors the occurrence of congestion and discards the tagged packet when it occurs. For example, all packets with the first tag are discarded at the time of congestion (such as a high bandwidth utilization rate). On the other hand, in normal times (such as when the bandwidth usage rate is low), the packet is discarded by referring not only to the first tag but also to the second tag.

  The congestion control device 1 illustrated in FIG. 1 includes a packet analysis unit 20, a malicious flow identification unit 22, a control order identification unit 24, and a communication control unit 26.

  The packet analysis unit 20 refers to the packet in the overflow queue and extracts a flow by bundling packets having the same flow identification information. As a result, by performing communication control on a flow basis, not on a packet basis, efficient communication control can be performed when constant regulation is taken.

  Then, the malignant flow specifying unit 22 compares the extracted flow with statistical information of the malignant flow input in advance, and determines whether the flow is a malignant flow. The statistical information is obtained by measuring characteristic parameters of the malignant flow such as the duration of the flow. Thus, by using the statistical information of the malignant flow, tagging that completely captures the characteristics of the malignant flow can be performed, and the deviation between the band design and the actual band use state can be adjusted.

  FIG. 5 is an explanatory diagram showing the control order specifying unit 24. When there are a plurality of malignant flows, the control order specifying unit 24 specifies the order in which the malignant flows are controlled. Examples of methods for determining the order in which the control order specifying unit 24 controls the malignant flow include, for example, a flow order in which the size is large, a dictionary sequence order (order from the smallest address number, time order in which information has arrived, etc.) The order calculated by the algorithm and the order determined from the information about the network collected by the control order specifying unit 24 can be mentioned.

  The communication control unit 26 controls communication of the identified malignant flow. The communication control is, for example, limiting the flow band to a predetermined value. As a result, communication control is performed while aiming at the malignant flow, so that the throughput of the majority of users is not reduced.

  With the congestion control system described above, it is possible to constantly realize congestion control without impairing the convenience of the user who is using it normally.

Note that the tagging process by the tagging device 14 is realized by any of the methods shown in the following table, for example.

  Method 1-A is a method for tagging a flow having a long flow duration to be preferentially discarded. As a result, communication control can be performed effectively when quality degradation over a long period of time occurs. A flow that has a characteristic of occupying a band for a long time may also adversely affect the network. You can capture and tag such flows reliably. In the conventional technique, tagging is performed with only a small number of parameters such as a flow rate and a burst size. Therefore, the duration is unknown and the malignant flow cannot be detected.

  Method 1-B is a method of tagging by changing the value according to the communication amount of the flow from the communication start time. As a result, it is possible to identify a flow having a high rate instantaneously and a flow having a constant high rate for a predetermined period of time for tagging, and to discard a packet with a priority. Therefore, the level of malignant flows can be accurately classified, and communication control can be effectively performed by narrowing down to flows that use a lot of bandwidth when congestion occurs.

  In method 1-B, the amount of flow traffic from the communication start time is set to {src, *} (a set of packets as a specific transmission source) or {*, dst} (a set of packets as a specific transmission destination) ) May be measured and tagged. Note that src is an abbreviation for transmission source (source), and dst is an abbreviation for transmission destination (destination).

  Further, the possible value (value range) of the tag value of the packet included in the tag may be two stages (see FIG. 6A) or may be a stage larger than two (see FIG. 6B). For example, for a packet having a two-stage value as the first tag and an N-stage value as the second tag, the priority discarding device 16 discards the packet by referring to only the first tag at the time of congestion. Sometimes the packet may be discarded by referring to both the first tag and the second tag (see FIG. 6C). For example, the packet P11 is discarded by referring to the value “1” of the first tag at the time of congestion, and by referring to the value “1” of the first tag and the value “0” of the second tag in normal times. pass. During this normal time, the packet P13 having the second tag value of 2 or more is discarded. As a result, when there is a difference between the design bandwidth and the actual usage status, it is possible to keep the user's throughput as high as possible by discarding only the minimum necessary packets when performing communication control.

  Furthermore, the tag value may be calculated again and again, not just once, based on information about the network such as the usage rate and the degree of congestion. Thereby, it is possible to realize communication control in accordance with the state of the network that keeps changing dynamically. Information about the network may be collected by measurement by an independent device such as a measurement device.

The packet discarding process by the priority discarding device 16 is realized by any of the methods shown in the following table, for example.

  Method 2-A is a method of determining a packet of a flow to be discarded by looking at a tag (see FIG. 7A). Packets with tag values of 1 (P2, P3, P4, P6) are discarded. As a result, when there is a margin in the network bandwidth, the throughput of the user can be kept as high as possible by narrowing down the target to only the malicious flow and discarding the packet.

  Method 2-B is a method of discarding all or part of the packets discarded by looking at the tag and the flow. All the packets (P2, P3, P6) of the flow (F2) having a tag value of 1 and a large size are discarded (see FIG. 7B). Alternatively, some packets (P2, P6) of the flow (F2) having a tag value of 1 and a large size are discarded (see FIG. 7C). As a result, when there is a difference between the design bandwidth and the actual usage status, it is possible to keep the user's throughput as high as possible by discarding only the minimum necessary packets when performing communication control.

The malignant flow specifying process by the malignant flow specifying unit 22 is realized by any of the methods shown in the following table, for example.

  Method 3-A is a method for identifying the malicious flow and its src address by counting the number of packets with the header information of the packets stored in the overflow queue. Thereby, since all the packets are counted, a malignant flow can be specified accurately.

  Method 3-B is a method in which method 3-A is performed on packets that have a tag with a tag value greater than or equal to a predetermined value among packets stored in the overflow queue. Thereby, referring to only the packets that are likely to be malignant flows with reference to the tag value, the number of packets to be counted can be reduced, and the malignant flows can be quickly identified.

  In scheme 3-A or scheme 3-B, header information includes a set of {src, dst}, {src, *}, {*, dst}. The src address includes an IP address, an MPLS label, an encapsulated address such as IPoverP, and the like. All referenced src addresses may be targeted, or only the top X src addresses with a large number of discarded packets may be targeted.

The communication control process by the communication control unit 26 is realized by any of the methods shown in the following table, for example.

  In method 4-A, while performing communication control, flow information is collected from the overflow queue of the priority discard apparatus 16 in which congestion has occurred, the congestion state is confirmed, and communication control is instructed in order until the congestion is alleviated. . If the congestion is not alleviated, issue an instruction again. Thus, by referring to the latest congestion state in the process of performing communication control of flows in order, it is possible to identify a malignant flow and to avoid communication control of a flow that is not a malignant flow.

  In the method 4-B, a plurality of flows are first subjected to communication control by a predetermined amount, and thereafter, an instruction is issued according to the order by the control order specifying unit 24 to perform communication control. The congestion control device 1 collects flow information from the priority discard device 16 in which congestion has occurred during that time, and issues instructions in order while confirming the congestion status. Thereby, congestion can be quickly alleviated by narrowing a plurality of flows by a predetermined amount.

  Method 4-C performs communication control by gradually controlling the amount of communication to be controlled in accordance with the congestion state. Communication control may be performed simultaneously or in the order determined by the control order specifying unit 24. Accordingly, since the flow rate is determined while checking the congestion state, it is possible to prevent communication control more than necessary, and to keep the user's throughput as high as possible.

  Method 4-D performs communication control step by step while monitoring a network congestion state by skipping monitoring packets such as ping when communication control is performed on a plurality of identified malicious flows. As a result, the flow rate suitable for mitigating congestion is known, and as a result, only the minimum communication control necessary for mitigating congestion is performed, avoiding excessively narrowing the bandwidth of malignant flows, and maximizing user throughput. It can be kept as high as possible.

  In the method 4-E, when communication control is performed on a plurality of identified malicious flows, the congestion control device 1 periodically measures the packet loss rate of the priority discard device 16, and calculates the flow rate to be controlled therefrom. Then, communication control of the malignant flow is performed at an appropriate flow rate. As a result, the flow rate suitable for mitigating congestion is known, and as a result, only the minimum communication control necessary for mitigating congestion is performed, avoiding excessively narrowing the bandwidth of malignant flows, and maximizing user throughput. It can be kept as high as possible.

  The present invention described above can be widely modified without departing from the spirit thereof as follows.

  For example, the communication control unit 26 needs to determine at which point communication control is performed in the path of the malignant flow from the transmission source to the transmission destination. About this, from the viewpoint of minimizing adverse effects on congestion due to the malignant flow, it is desirable that the point be as close as possible to the source (transmission source) of the malignant flow. Therefore, the communication control unit 26 accesses network topology information managed by a routing protocol or the like, and identifies the source of the malicious flow.

  Here, when going back to the source of the malicious flow, the network management company is different and the address system between networks is different, so if you do not know where to go after looking at the address of the malicious flow It is desirable to arrange an agent having an address conversion means between carrier networks to convert the address system.

  Further, the order determination process (S26) for controlling the malignant flow may be omitted. As an example that needs to be omitted, when only one malignant flow is found, there is a case where an urgent recovery is required so that a recovery time for controlling the malignant flows one by one cannot be obtained. When the order determination is omitted, communication control is performed simultaneously for a plurality of malignant flows (S30). As a result, it is possible to shorten the time required for network recovery by congestion control.

  Further, the congestion control device 1 is configured to accommodate each component (packet analysis unit 20, malignant flow identification unit 22, control order identification unit 24, communication control unit 26) in one housing. Is merely an example, and the number of cases constituting the congestion control device 1 may be configured as a plurality, and each component may be distributed. Each component of the congestion control device 1 may be in the same network or in a different network. If the networks are different, they may be the same operator or different operators. If each network has different functions, there is an agent that has an address translation function between them. As a result, the CPUs of the plurality of congestion control devices 1 perform calculations, respectively, so that load distribution can be realized.

  Moreover, you may accommodate each component of the congestion control apparatus 1, and the apparatus which has another function in the same housing | casing. For example, the communication control unit 26 may be built in the communication device 12, and message passing for communication control processing may be realized not on a network line outside the housing but on an internal bus inside the housing. Thereby, high-speed communication of the internal bus and load avoidance of the network line can be realized.

  In this specification, the congestion control device 1 is not a communication device 12, a tagging device 14, and a priority discard device 16, but is configured as an independent device on the network as shown in FIG. A configuration in which at least one component of the congestion control device is built in the communication device 12 is referred to as a distributed type.

  Furthermore, although the congestion control device 1 in FIG. 1 is configured to house each component in one housing, this configuration is merely an example, and a plurality of housings constituting the congestion control device 1 can be provided. The components may be distributed and arranged. Each component of the congestion control device 1 may be in the same network or in a different network. If the networks are different, they may be the same operator or different operators. If each network has different functions, there is an agent that has an address translation function between them. As a result, the CPUs of the plurality of congestion control devices 1 perform calculations, respectively, so that load distribution can be realized.

  Moreover, you may accommodate each component of the congestion control apparatus 1, and the apparatus which has another function in the same housing | casing. For example, the communication control unit 26 may be built in the communication device 12 and the message passing of the communication control process may be realized not on the network line outside the housing but on the internal bus inside the housing. Thereby, high-speed communication of the internal bus and load avoidance of the network line can be realized.

  Furthermore, each component of the congestion control device 1 may be realized by the CPU executing a program stored in a dedicated device of each component, or stored in a functional block of the communication device 12. It may be realized by executing the programmed program.

  Further, a plurality of components of the congestion control device 1 may exist by starting a plurality of processes. The communication device 12 in charge is assigned to each of the plurality of components. As a result, even when there are many communication devices 12, data parallel processing between a plurality of components can be realized, the time required to suppress congestion can be shortened, and the number of communication devices that perform communication as processing partners can be reduced. Therefore, there is an advantage that scalability with respect to the number of devices is improved and management is easy.

  Here, the correspondence information between each component of the congestion control device 1 and the communication device 12 in charge may be created dynamically or statically. As an example of creating dynamically, there is a method of preferentially allocating the communication device 12 to a component having a current calculation load smaller than other components among a plurality of components having the same function. On the other hand, as an example of static creation, there is a method of referring to a database that stores correspondence information between components created by an administrator in advance and the communication device 12.

  Further, the format (line type, communication protocol, etc.) of the control target network 10 is not limited to a specific one. For example, the network 10 to be controlled transfers a circuit-switched network such as TDM (Time Division Multiplexing) and WDM (Wavelength Division Multiplexing) or IP (Internet Protocol), Ethernet (registered trademark), and IP packets with labels. It may be an MPLS network. Each communication device is connected by a logical path provided by setting a lower layer path positioned relatively lower than the network layer in which the node exists. The communication device 12 may be, for example, an MPLS router of an MPLS network or a router of a packet network.

It is a block diagram which shows the congestion control system regarding one Embodiment of this invention. It is explanatory drawing which shows the tagging process by the tagging apparatus regarding one Embodiment of this invention. FIG. 3 is a block diagram illustrating a tagged packet according to an embodiment of the present invention. It is explanatory drawing which shows the priority disposal apparatus regarding one Embodiment of this invention. It is explanatory drawing which shows the control order specific | specification part regarding one Embodiment of this invention. It is explanatory drawing which shows the tag value of the packet regarding one Embodiment of this invention. It is explanatory drawing which shows the discard process of the packet regarding one Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Congestion control apparatus 10 Control object network 14 Tagging apparatus 16 Priority discard apparatus 20 Packet analysis part 22 Malignant flow specification part 24 Control order specification part 26 Communication control part

Claims (16)

A tagging device for tagging a packet of a flow to be relayed, a discarding device for discarding the packet in an overflow queue with reference to the tag, and congestion generated in a network configured to include a communication device for relaying the flow, A congestion control method controlled by a congestion control device,
The congestion control device has a packet analysis unit, a malicious flow identification unit, a communication control unit,
The packet analysis unit extracts a flow from the overflow queue packet according to flow identification information;
The malignant flow identification unit collates the extracted flow feature information with the malignant flow feature information stored in the storage means, and identifies whether the extracted flow is a malignant flow; ,
The communication control unit performs communication control of the identified malignant flow;
The congestion control method characterized by performing. The congestion control device has a control order specifying unit,
The congestion control method according to claim 1, wherein the control order specifying unit executes a procedure for defining an order of malignant flows controlled by the communication control unit.   The congestion control method according to claim 1, wherein the tagging device performs tagging so as to preferentially discard a flow having a long flow duration.   The congestion control method according to claim 1, wherein the tagging device performs tagging by changing a value according to a communication amount of a flow from a communication start time.   The congestion control method according to claim 1, wherein the tagging device measures the traffic of the flow for each packet that is a specific destination or a specific source.   6. The congestion control method according to claim 1, wherein the discard device discards a packet in an overflow queue with reference to characteristic information of a flow to which the tag and the packet belong.   The said malicious flow specific | specification part counts the number of packets based on the header information of the packet stored in the said overflow queue, The characteristic information of the extracted flow is acquired, The Claim 1 thru | or Claim characterized by the above-mentioned. 6. The congestion control method according to any one of 5 above.   2. The malicious flow identification unit counts the number of packets based on packet header information and a tag value stored in the overflow queue, and acquires extracted flow feature information. The congestion control method according to any one of claims 5 to 5.   9. The congestion control method according to claim 2, wherein the communication control unit repeats confirmation of a congestion state and communication control until congestion is alleviated.   9. The congestion according to claim 2, wherein the communication control unit performs communication control for each of the flows according to an order after performing communication control of a plurality of flows by a predetermined amount. Control method.   The congestion control method according to any one of claims 2 to 8, wherein the communication control unit performs communication control in a stepwise manner according to a congestion state.   The congestion control method according to any one of claims 2 to 8, wherein the communication control unit acquires a congestion state using a monitoring packet.   The congestion control method according to any one of claims 2 to 8, wherein the communication control unit acquires a congestion state based on a packet loss rate of the communication apparatus. Controls congestion occurring in a network including a tagging device for tagging a packet of a flow to be relayed, a discarding device for discarding the packet in an overflow queue with reference to the tag, and a communication device for relaying the flow A congestion control device,
A packet analysis unit for extracting a flow from packets in the overflow queue according to flow identification information;
A malignant flow identifying unit that collates the extracted flow feature information with the malignant flow feature information stored in the storage unit, and identifies whether the extracted flow is a malignant flow,
A communication control unit for controlling communication of the identified malignant flow;
A congestion control apparatus comprising: A tagging device that tags a packet, a discard device that discards the packet in an overflow queue with reference to the tag, a communication device that relays a flow, and a flow in accordance with the flow identification information is extracted from the packet in the overflow queue, The feature information of the extracted flow and the feature information of the malignant flow stored in the storage unit are collated to identify whether or not the extracted flow is a malignant flow, and the identified malignant flow is communicated The tagging device used in a congestion control system configured to include a congestion control device to control,
A tagging device that tags a packet of a flow to be relayed. A tagging device for tagging a packet of a flow to be relayed, a discarding device having an overflow queue, a communication device for relaying a flow, and a flow from the packet of the overflow queue according to flow identification information, and the extracted flow The congestion control device that collates the feature information of the malignant flow stored in the storage means, identifies whether the extracted flow is a malignant flow, and performs communication control of the identified malignant flow A discarding device used in a congestion control system comprising:
A discarding device that discards a packet to an overflow queue by referring to the tag of the packet.

Images