JP2005150955A - Path controller, communication controller, and communication system employing them - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To realize flow control at a network core part with a minimum processing load while making the best use of the advantages of the class control at the network core part and the control in flow units at edge parts. <P>SOLUTION: A communication network includes: a communication controller (10) for managing communication on the network; ingress routers (20) through which packets transmitted to the network pass; and control routers (30) for carrying out flow control of the packets inside the network. The communication controller particularizes the ingress router and the control router for each packet flow, and transmits a marking request for designating a concerned control router and the flow control carried out by the concerned control router to the particularized ingress router. The ingress router attaches the marking information including the control router and the flow control to the packets included in the packet flow according to the marking request. The control router controls the packet flow on the basis of the marking information of the packets. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a packet communication system, and in particular, enables efficient flow control in a core portion using packet marking, assuming class control in a core portion of a network and individual flow control in an edge portion. The present invention relates to a communication system, and a communication control device and a router used in such a communication system.

  Conventionally, communication via the Internet has been developed by treating all packets as the same best effort type. The best effort type is a communication form in which the entire network strives to do its best, but there is no guarantee of service from the end-to-end, and a certain service type is not guaranteed.

  On the other hand, in recent years, a control method for performing QoS control on a network such as Diff-serv and Int-serv has been proposed (see, for example, Non-Patent Documents 1 and 2). Diff-serv is a technology aiming to secure QoS throughout the network based on the concept of class. Specifically, each router in the network has a class (DSCP: Diffserv Code) set in the packet. By performing control based on Pointn), a certain level of quality as a whole network is ensured for each class. On the other hand, Int-serv is a technology that secures QoS for each flow by installing a path for each flow between routers in a network.

  Diff-serv has a merit that the control load of the router is light because the control is performed in units of class, and the scalability for a large-scale network is high. However, since priority control is performed in units of classes, there is a demerit that it is difficult as a real problem to secure QoS in units of flows.

  On the other hand, Int-serv has an advantage that QoS can be easily secured in units of flows. However, since the RSVP protocol (resource reservation protocol) for realizing Int-serv is complicated, the processing load on the router is heavy and large. There is a demerit that the scalability for a large-scale network is low. (Refer nonpatent literature 2).

By combining the merits of both Diff-serv and Int-serv, QoS control is performed in units of flows in Int-serv at the edge of the network, and Diff-serv is controlled inside the network (core part). Attempts have been made to build highly scalable networks. As an example, based on the Diff-serv control method, a method of marking a class that differentiates a service in a packet at an edge portion has been proposed (see, for example, Patent Document 3).
“An Architecture for Differentiated Service”, IETF (Internet Engineering Task Force), RFC 2475 "" Resource Reservation Protocol "IETF (Internet Engineering Task Force), RFC2205 JP 2002-111742 A

  However, by simply combining the conventional Diff-serv system and Int-serv system, it is possible to cope with flow control required in an actual mobile communication network, such as buffering processing at the time of handover of a mobile terminal, at the core part. Have difficulty. As shown in FIG. 1, flows having the same priority among flows passing between routers 1 and 2 in the network are treated as one class. In the example of FIG. 100, the flows A1 to A4 are treated as the same class, the flows B1 and B2 are treated as the same class, and the flows C1 and C2 are treated as the same class, and are collectively controlled without identifying individual flows. Here, when the mobile terminal requesting the flow A3 enters the handover, the router 2 checks all the flows sequentially in order to buffer the packet of the control target flow A3 and route it to the destination. There must be. That is, the flow to be controlled must be identified, and the control content (control method) required by each flow must be determined before entering the flow control operation. This increases the burden on the router at the core and impairs the merits of class control.

  Although the technique disclosed in Patent Document 3 also performs marking at the edge portion, there is no difference from the conventional Diff-serv control as a whole, and there is no mention of QoS control focusing on individual flows at the core portion. Not done.

  In this way, although a method for performing flow unit control at the edge part and class control at the core part has been proposed, the core part such as buffer control at handover and packet arrival time designation control in the mobile communication network has been proposed. Considering the special control in the core, the necessity of control of the flow unit in the core cannot be ignored. Therefore, an improved communication system that efficiently performs the flow control in the core portion with a low load while taking advantage of the flow control in the edge portion and the class control in the core portion is desired.

  Therefore, the present invention presupposes Diff-serv class control in the core portion and Int-serv individual flow control in the edge portion, and a plurality of flows for each control content (control method) required in the core portion. Are treated as the same group. The group here is different from the class based on the Diff-serv packet priority, and is based on the Diff-serv class control and is used for various control such as buffering control and timed delivery control. This is a flow group according to the content (method). Furthermore, the individual flows in the group are efficiently controlled while handling the same flow group collectively.

  In order to achieve this, an identifier indicating the control router through which each flow passes and the control content to be performed at the ingress router to the core part (that is, the router at the edge part) is used by using a marking method. Mark. The control router inside the core looks only at the marking to determine whether it needs to be controlled by its own router, and if control is required, it simply performs the control specified by the marking without identifying the flow individually. It's okay. Thereby, it is possible to realize control of individual flows while minimizing the burden on the router for flow control in the core portion.

  Specifically, in the first aspect of the present invention, a communication control device that manages communication on a network, an ingress router through which a packet transmitted into the network passes, and flow control of the packet inside the network A communication system including a control router that performs In this communication system, the communication control device specifies an ingress router and a control router for each packet flow, and specifies a corresponding control router and a flow control performed by the control router for the ingress router. Send. The ingress router adds marking information including the control router and flow control to the packet included in the packet flow according to the marking request. The control router controls the packet flow based on the marking information of the packet.

  According to such a communication system, it is possible to perform flow control in the network core unit while suppressing the load on the control router.

In the 2nd side surface of this invention, the communication control apparatus used with the communication system mentioned above is provided. The routing device
(A) an execution management unit that detects the necessity of flow control for packets transmitted and received on the network;
(B) For a packet flow that needs flow control, a flow management unit that identifies an ingress router through which the packet flow passes and a control router that performs flow control on the packet flow;
(C) The control management part which transmits the marking request | requirement which designates a corresponding control router and the flow control performed by this control router with respect to the specified entrance router is provided.

  By using such a communication control device, it is possible to realize marking processing of information necessary for flow control at the edge portion of the network for a packet flow that requires flow control inside the network.

According to a third aspect of the present invention, there is provided a path control device as an ingress router that is used in the above-described communication system and through which a packet transmitted in a network passes. The routing device
(A) Control management for receiving a marking request including a control identifier representing a flow control necessary for the packet and a router identifier representing a control router on which the flow control is performed, from a communication control device that manages communication on the network. And
(B) an execution management unit that identifies a flow to which a packet that has arrived at the path control device belongs, and determines whether a marking request for the flow has been received;
(C) a marking control unit that adds marking information including the control identifier and the router identifier to the arriving packet when the marking request is received.

  By using such a path control device, it is possible to mark a control router that performs flow control and a flow control method performed therefor for each packet flow that requires flow control in the network. As a result, flow control within the network is performed simply and with low load.

According to a fourth aspect of the present invention, there is provided a path control device as a control router that controls the flow of packets passing through a network. The routing device
(A) an execution management unit that determines whether an arrival packet includes marking information that specifies flow control in the route control device;
(B) a packet control unit that controls the arrival packet according to the marking information when the arrival packet includes marking information for designating flow control in the route control device; By using such a path control device, the designated flow control can be performed quickly and accurately only by checking the marking information of the arrival packet at the network core unit.

  While taking advantage of the QoS architecture of flow control at the edge part and class control at the core part, it is possible to realize flow control as necessary at the core part with a minimum burden on the router.

  Other features and effects of the present invention will become more apparent from the embodiments described below with reference to the drawings.

  FIG. 2 is a schematic diagram of a communication system 1 to which the present invention is applied. As shown in FIG. 2A, a communication system 1 includes a communication control device 10 that controls communication on a network, edge routers (ER) 20a to 20d that perform flow control in an edge region of the network, and a network core. Control routers (CR) 30a to 30c that perform flow group control in the area are included. The edge router 20 individually controls the flow for each communication terminal using this network. On the other hand, as shown in FIG. 2B, the control router 30 in the core area divides the flow into groups for each control content (control method) required by the flow, and handles the flow in units of groups. For example, group A is a group that requires buffering control, group B is a group that requires timed distribution, and group C is a group that does not require special control and only requires normal communication control. The grouping of flows according to such control contents is performed in parallel with the Diff-serv class control. Therefore, even flows belonging to the same priority class in Diff-serv may be divided into different groups depending on the control method required by the flow.

  The communication control apparatus 10 manages any control required on the network, such as a call request, a mobile terminal handover, a time-designated distribution request, a retransmission request, and a packet count request. Then, the edge router 20 at the entrance where each packet flow passes and the control router 30 on which flow control is to be performed are specified. The communication control device 10 determines what control contents (control method) are required by which control router 30 for each packet flow, and designates the control router identifier representing the specified control router 30 to the corresponding edge router 20. And requesting to mark a control identifier representing the control content.

  Upon receipt of the marking request, the edge router (ER 20a in the example of FIG. 2A) marks the control router and an identifier indicating the control content on each packet of the flow, and sends the packet to the control router 30 in the core area. The control router 30 determines the control router identifier of each flow that has been sent, determines whether it is addressed to its own router, and if it is addressed to its own router, it performs flow control according to the control identifier. Just do it.

  With such a configuration, individual control of flows can be performed without increasing the load on the control router 30 in the core area of the network.

  FIG. 3 is a configuration diagram of the communication system 1A according to the first embodiment of the present invention. In the first embodiment, flow control will be described by taking buffering control at the time of handover of a mobile terminal as an example.

  The mobile communication network 50 of the communication system 1A includes a communication control device 10, an entrance router 20, a control router 30, base stations 40A and 40B, and routers 31 to 33. The mobile terminal 12 is a mobile phone used directly by the user, for example, and communicates with the counterpart communication device 11 via the network 50. The counterpart communication device 11 may be a terminal device or a server. In the example of FIG. 3, the counterpart communication device 11 is connected to the network 50 directly or via another network, but may be included in the network 50.

  In FIG. 3, for convenience of explanation, a single communication control device 10, an ingress router 20, and a control router 30 are shown, but the network 50 includes a plurality of communication control devices 10, an ingress router 20, and a control router 30. The routers 31 to 33 may function as control routers or entry routers. In this case, the plurality of communication control devices 10 are connected so as to cover all routers, entry routers, and control routers on the network 50, and control and manage communication on the network 50.

  The base station 40A is connected to the control router 30 via the router 32 by a signal line. The base station 40B is connected to the control router 30 via the router 33 by a signal line. The communication control device 10, the entrance router 20, and the control router 30 are connected to each other by a signal line. The mobile terminal 12 is moving from the service area of the base station 40A to the service area of the base station 40B. The communication control device 10 detects a handover of the mobile terminal 12, supplies control information to the ingress router 20 and the control router 30, and performs operations according to the control information at the ingress router 20 and the control router 30, thereby Efficient flow control in 50 core parts is possible. That is, the communication control device 10, the ingress router 20, and the control router 30 constitute the communication system of the present invention and perform service control in cooperation.

  FIG. 4 is a sequence diagram showing a processing flow of the entire communication system 1A shown in FIG. As described above, in the first embodiment, the present invention is applied to flow control triggered by the handover of the mobile terminal 12.

  First, while the activation terminal 12 is located in the service area of the base station 40A, a packet is sent from the counterpart communication device 11 to the base station 40A via the ingress router 20 and the control router 30 by normal routing. The mobile terminal 12 receives a packet from the base station 40A. The mobile terminal 12 periodically transmits position information to the communication control apparatus 10, and the communication control apparatus 10 registers the position of the mobile terminal 12. In the example of FIG. 4, the position information is transmitted from the mobile terminal 12 to the communication control apparatus 10. However, the present invention is not limited to this example, and the position of the mobile terminal 12 is transferred from the base station 40 </ b> A to the communication control apparatus 10. Information may be supplied, or the position information may be sent to the communication control device 10 from an intermediate exchange or control device (both not shown).

  Here, when the mobile terminal 12 starts moving from the service area of the base station 40A to the service area of the base station 40B, the communication control apparatus 10 detects a handover of the mobile terminal 12 (S101). The communication control device 10 determines the ingress router 20 and the control router 30 through which the packet flow from the counterpart communication device 11 passes (S103), and sends a marking request (S105) and a control start request (S107) to each. send. Upon receiving the marking request, the ingress router 20 marks the control router identifier and the control content identifier on the packet from the counterpart communication device 11 to the mobile terminal 12 (S200), and transfers the marked packet to the control router 30. The control router 30 sees only the identifier of the own router marked in the packet and the identifier of the buffering process, and starts buffering of this packet flow (S300).

  When the mobile terminal 12 completes the movement of the base station 40B into the service area, the communication control device 10 detects the end of handover of the mobile terminal 12 (S109). Then, a marking release request is sent to the ingress router 20 (S111), and a control end request is sent to the control router 30 (S113).

  The control router 30 releases the buffer according to the control end request, and transmits the packet buffered during the handover to the destination base station 40B (S400). In response to the marking cancellation request, the ingress router 20 does not perform marking on the packet flow from the counterpart communication device 11 to the mobile terminal 12 thereafter. Therefore, the subsequent packet flow is sent to the base station 40B via the ingress router 20 and the control router 30 by a normal routing process.

  FIG. 5 is a schematic configuration diagram of the communication control apparatus 10, and FIG. 6 is a flowchart showing processing performed in the communication control apparatus 10. FIG. 6 shows the operation of the communication control apparatus 10 in FIG. 4 in detail, and the same steps as those in the sequence diagram of FIG. 4 are denoted by the same reference numerals.

  The communication control apparatus 10 includes a control management unit 121, an execution management unit 122, and a flow management unit 123. As shown in the flowchart of FIG. 6, first, the execution management unit 122 of the communication control apparatus 10 detects a handover of the mobile terminal 12 in the network 50 based on the location information of the mobile terminal 12 (S101). The execution management unit 122 that has detected the handover makes an inquiry to the flow management unit 123, and the ingress router 20 through which the packet flow sent from the counterpart communication device 11 to the mobile terminal 12 passes, and the control that should perform the buffering control accompanying the handover. The router 30 is specified (S103).

  When the ingress router 20 and the control router 30 are identified, the execution management unit 122 of the communication control apparatus 10 sends a marking request to the identified ingress router 20 via the control management unit 121 (S105). In the case of the first embodiment, a control start request is sent to the control router 30 together with a marking request to the ingress router 20 (S107). This is because the control router 30 needs to secure resources for buffering accompanying the handover.

  The marking request to the ingress router 20 includes, as notification information, a flow identifier of a packet flow sent from the counterpart communication device 11 to the mobile terminal 12 and marking information. The marking information includes a control identifier indicating the control content (buffering control in the first embodiment) required by this packet flow, and a control router identifier indicating the control router 30 that executes control. On the other hand, the control start request to the control router 30 includes the flow identifier of the packet sent to the mobile terminal 12. When receiving the control start request, the control router 30 reserves a buffer area for the flow specified by this flow identifier.

  If the movement of the mobile terminal 12 is completed, the execution management unit 122 of the communication control apparatus 10 detects the end of the handover (S109). The execution management unit 122 of the communication control device 10 issues a marking release request to the ingress router 20 via the control management unit 121 (S111). A control end request is issued to the control router 30 (S113). The notification information attached to the marking release request includes the flow identifier, but the marking information may be attached as it is. Further, the notification information attached to the control end request includes a flow identifier. The control router 30 releases the control for the flow for which the control end request has been issued. In the case of the first embodiment, the control router 30 releases the buffer area allocated to the flow.

  Such a communication control device 10 is realized by, for example, a control unit of an exchange.

  FIG. 7 is a schematic configuration diagram of the ingress router 20 used in the communication system 1A of FIG. 3, and FIG. 8 is a flowchart of processing operations performed in the ingress router 20. The processing flow of FIG. 8 is details of the marking (S200) processing in the sequence of FIG.

  The ingress router 30 includes a control management unit 101 connected to the communication control device 10, an execution management unit 102, a flow management unit 103, an input control unit 104 connected directly or indirectly to the counterpart communication device 11, and a next router 31. An output control unit 106 connected to the input control unit 104, and a marking control unit 105 that performs marking on a packet that has arrived at the input control unit 104. The flow management unit 103 receives the marking request sent from the communication control device 10 via the control management unit 101, and writes the marking information (see FIG. 6) included in the marking request in the flow management table 25.

  In operation, when a packet arrives at the input control unit 104, the execution management unit 102 identifies the flow to which the arrived packet belongs (S201). Then, the flow management table 25 stored in the flow management unit 103 is referred to by the notification from the communication control apparatus 10 (S203), and it is determined whether or not the marking process is necessary for this packet (S205). If the marking process is unnecessary (NO in S205), the packet is sent from the output control unit 106 to the next router 31 as it is (S209). If marking processing is necessary (YES in S205), the marking control unit 105 adds the marking information specified in the flow management table 25 to the packet (S207), and then sends the packet to the output control unit 106.

  As described above, the ingress router 20 performs marking when marking is required for a packet passing through the ingress router 20 based on the marking information sent from the communication control device 10. In the case of the first embodiment, the ingress router 20 starts marking a packet sent from the counterpart communication device 11 to the mobile terminal 12 in response to a marking request from the communication control device 10. Although not shown, it is assumed that QoS class control is performed as usual in parallel with the marking process.

  FIG. 9 shows an example of the flow management table 25 of the ingress router 20. The flow management table 25A illustrated in FIG. 9 stores a flow identifier and marking information in association with each other, and the marking information includes a router identifier and a control identifier. Although flow # 1 and flow # 205 are separate flows, both require buffering control in the control router 30. Flows # 21, # 25, and # 54 require time designation control in the control router 30. The flow # 72 requires multicast control from the routers 31 and 32, the flow # 93 requires a packet count at the router 33, and the flow # 98 requires a packet count at the router 32.

  Here, the flow of the same control content, that is, the flow # 1 and the flow # 205 that require buffering are treated as the same group between the routers of the core unit. Also, flow # 93 and flow # 98 that require packet count are handled as the same group.

  Depending on the contents of the control, there are cases where individual control is required even in the flows grouped in the same group as the same control method. For example, in the case of buffering control accompanying handover as in the first embodiment, a separate buffer area must be secured for each flow in the control router 30. Therefore, in the flow management table 25A of FIG. 9, information for distinguishing each flow is included in the control identifier itself so that control for different flows in the same group is distinguished. Specifically, the flows are distinguished within the same control method, such as packet number count # 1 and packet number count # 2.

  The form of the control identifier included in the marking information can be determined by the communication control device 10. In the case of the first embodiment, for example, when the communication control apparatus 10 detects a handover with respect to the packet flow # 205, the router identifier of the control router 30 and the control of buffering control # 2 are included in the marking request sent to the ingress router 20. Include an identifier. The flow management unit 103 of the ingress router 20 writes this notification information in the management table 25A.

  Instead of the flow management table 25A of FIG. 9, a flow management table 25B shown in FIG. 10 may be used. In the flow management table 25B, the flow identifier is included in the marking information. In the case of FIG. 10 as well, as in FIG. 9, flows that require the same control content are treated as the same group, but the router and the control content are recorded in association with the flow identifier in the marking information. Only the same control identifier is given to the control contents.

  FIG. 11 shows still another flow management table 25C. In the flow management table 25C of FIG. 11, the flow identifier is not included in the marking information, and the control identifier does not have information for distinguishing the flows.

  As described above, the form of the marking information sent from the communication control unit 10 to the ingress router 20 can be designated by the communication control apparatus 10. Depending on the configuration of the marking information generated by the communication control device 10, the ingress router 20 may store any of the flow management tables 25A, 25B, and 25C.

  FIG. 12 shows a processing flow when the ingress router 20 receives a marking-related request from the communication control device 10. The ingress router 20 updates the flow management table 25 when it receives a marking request (S211) and when it receives a marking release request (S213). For example, in the example of the flow management table 25A, when the marking request for the flow # 1 is received, the router identifier (control router 30) and the control identifier (buffering # 1) of the marking information are associated with the flow identifier of this flow. Write to table 25A. Even when a marking request for another flow is received, data is sequentially written in the table 25A.

  When the ingress router 20 receives the marking release request for the flow # 1 from the communication control device 10, the ingress router 20 deletes the entry of the flow # 1 from the flow management table 25A. Therefore, after this, the packet is transferred to the next router by a normal procedure without marking the packet included in this flow.

  FIG. 13 is a schematic configuration diagram of the control router 30 used in the communication system 1A of FIG. 3, and FIG. 14 is a flowchart of processing operations performed in the control router 30. The processing flow of FIG. 14 shows details of processing corresponding to buffering (S300) in the sequence of FIG.

  The control router 30 includes a control management unit 111, an execution management unit 112, a flow management unit 113, and an input control unit 115 that are logically connected to the counterpart communication device 11 and process incoming packets. And an output control unit 116 connected to the next router, and a packet control unit 114 that performs predetermined control on packets belonging to the control target flow.

  First, the execution management unit 112 checks the presence or absence of marking of a packet that has arrived at the input control unit 115 (S301). The unmarked packet (NO in S301) is transferred to the next router as it is by normal routing (S311). If the packet is marked (YES in S301), the control router identifier included in the marking information is checked to determine whether it is addressed to the own router (S303). A packet that does not require control by the own router (NO in S303) is transferred to the next router as it is (S311). When the arrival packet is included in the control target flow of the own router (YES in S303), the control content required by the packet is identified from the information marked in the packet (S305), and the packet is determined according to the designated control method. Is controlled (S307).

  Until there is a control end request from the communication control device (NO in S309), the control on the arrival packet is continued. In the first embodiment, buffering to a specific buffer area is continued until there is a control end request. When the control end request is received (YES in S309), the packet is transferred to the next router (S311).

  In step S305, when the ingress router 20 uses the flow management table 25A shown in FIG. 9 or the flow management table 25B shown in FIG. 10, the execution management unit 112 of the control router 30 only sees the marking information stamped on the packet. Thus, the control method of the arrival packet can be identified for each flow. That is, only by checking the control identifier marked in the received packet (in the case of FIG. 9), or by simply checking the correspondence between the control identifier marked in the packet and the flow identifier (in the case of FIG. 10), the packet Can be stored in the buffer area reserved for the flow.

  On the other hand, when the ingress router 20 uses the flow management table 25C shown in FIG. 11, the marking information marked on the packet includes a router identifier and a control identifier with no flow distinction. In this case, the flow management unit 113 of the control router 30 has a control management table 35 shown in FIG. The execution management unit 112 of the control router 30 cannot determine in which buffer area the arrived packet should be buffered only from the marking information. Therefore, in step S305, the execution management unit 112 refers to the control management table of the flow management unit 113 as necessary to identify the control method of the arrival packet. When packet control is identified, packet control (buffering control in the first embodiment) is started, and control is continued until a control end request is received. After receiving the control end request, the buffer is released and the packet is transmitted to the next router.

  FIG. 15 shows an example of the control management table 35 held by the control router 30. When the control management table 35 is used, the control start request (see FIG. 6) sent from the communication control device 10 includes information on the control content (method) in addition to the flow identifier. When receiving the control start request, the control management unit 111 of the control router 30 records the flow identifier and the control identifier in the control management table 35 of the flow management unit 113 via the execution management unit 112. The execution management unit 112 refers to the control management table 35, confirms the packet flow corresponding to the control identifier of the received packet, and stores this packet in the corresponding buffer area.

  FIG. 16 shows processing when the control router 30 receives a control-related request from the communication control device 10. When the communication control apparatus 10 detects a handover of the mobile terminal 12, the control router 30 receives a control start request for the packet flow sent to the mobile terminal 12 from the communication control apparatus 10 (S401). The control router 30 secures a buffer area for this packet flow (S403). When the control management table 35 is used, the control management table 35 is updated along with securing the buffer area.

  On the other hand, when the communication control apparatus 10 detects the completion of movement of the mobile terminal, the control router 30 receives a control end request for the packet flow sent to the mobile terminal 12 (S405). Then, the buffer area is released according to the control end request, and the stored packets are transmitted to the base station that is the destination of the mobile terminal 12 (S407). When the control management table 35 is used, the entry for this packet flow is deleted from the control management table 35.

  In this way, the control router 30 does not need to identify all individual flows, and determines whether normal routing or specific control is performed according to the presence or absence of the marking information of the received packet. When the received packet has marking information and designates control in the own router, it is only necessary to perform control according to the control identifier included in the marking information without identifying the flow. As a result, group processing for the same control method is performed in the core portion, while individual flows in the group can be controlled without increasing the burden on the control router 30.

  Next, a second embodiment of the present invention will be described with reference to FIGS. FIG. 17 is a schematic configuration diagram of a communication system 1B according to the second embodiment of the present invention, and FIG. 18 is a flowchart showing the operation of the communication control 10 used in the communication system 1B.

  In the second embodiment, the present invention is applied to time designation distribution to a plurality of mobile terminals 12A to 12C. The time-designated distribution is necessary, for example, in the distribution of a new music with a designated date and time, or in an auction or quiz performed by specifying a time. This is because the information needs to be distributed at the same time so as to be fair to all users who wish to obtain new music, all users who participate in auctions, quizzes, and the like.

  The communication system 1B includes a communication control device 10, an ingress router 20, and a control router 30, similarly to the communication system 1A of the first embodiment. A router 31 is inserted between the ingress router 20 and the control router 30, and the control router 30 is connected to the base station 40A via the router 32. These are connected by a signal line, for example. The mobile terminals 12A to 12C are mobile terminals used directly by the user. The counterpart communication device 11 is a communication counterpart of the mobile terminals 12A to 12C, and may be a terminal or a server depending on the communication form, and may be included in the mobile communication network 50 and directly connected to the mobile communication network 50. Or connected via other networks.

  The communication control apparatus 10 includes a control management unit 121, an execution management unit 122, and a flow management unit 123, as in the first embodiment illustrated in FIG. In the flowchart of FIG. 17, the execution management unit 122 of the communication control device 10 is a flow in which the packet flow to the mobile terminals 12 </ b> A, 12 </ b> B, 12 </ b> C is subject to arrival time designation control through negotiation with the counterpart communication device 11, for example. (S501). The execution management unit 122 that has detected the target flow of the arrival time designation control makes an inquiry to the flow management unit 123 and arrives at the entrance router 20 through which the packet flow sent from the counterpart communication device 11 to the mobile terminals 12A, 12B, and 12C passes. The control router 30 that should perform the time designation control is specified (S503).

  In FIG. 17, for convenience of explanation, information is distributed to the mobile terminals 12A to 12C in the service area of the base station 40A by specifying the arrival time, but the arrival time to the mobile terminal in the service area of another base station Naturally, the specified distribution is also included. The communication control device 10 can specify the corresponding control router for each flow according to the position of the mobile terminal that is the distribution destination.

  When the ingress router 20 and the control router 30 are identified, the execution management unit 122 of the communication control apparatus 10 sends a marking request to the identified ingress router 20 via the control management unit 121 (S505). The marking request includes, as notification information, a flow identifier of a packet flow sent from the counterpart communication device 11 to the mobile terminal 12A, 12B, 12C, and marking information. The marking information includes a control identifier indicating the control content (arrival time designation delivery control in the second embodiment) required by the packet flow, and a control router identifier indicating the control router 30 that executes the control. In the second embodiment, arrival time information is also included in the control identifier.

  The execution management unit 122 of the communication control apparatus 10 detects the completion of transmission of a packet that needs to specify an arrival time through negotiation with the partner distribution apparatus 11 (S507). The execution management unit 122 sends a marking release request to the ingress router 20 via the control management unit 121 (S509). The notification information attached to the marking release request includes the flow identifier, but the marking information may be attached as it is. The communication control device 10 is realized by, for example, a control unit of an exchange.

  As in the first embodiment shown in FIG. 7, the ingress router 30 is directly connected to the control management unit 101, the execution management unit 102, the flow management unit 103, and the counterpart communication device 11 connected to the communication management device 10. Alternatively, it includes an input control unit 104 that indirectly receives a packet, a marking control unit 105 that performs marking control on the packet, and an output control unit 106 that controls output to the next router.

  The flow management unit 103 stores any of the flow management tables 25A to 25C of FIGS. When the packet arrives at the input control unit 104, the execution management unit 102 identifies the flow to which the arrived packet belongs, refers to the flow management table 25 stored in the flow management unit 103, and performs the marking process for this packet. Determine no. For example, assume that the flow management table 25A shown in FIG. 9 is used.

  In the flow management table 25A of FIG. 9, flow # 21, flow # 25, and flow # 54 are packet flows that should be delivered to the mobile terminals 12A, 12B, and 12C by specifying arrival times. These flows need to be subjected to arrival time designation control by the control router 30. Each time a packet arrives, the ingress router 20 refers to the flow management table 25B to determine whether or not the flow to which the packet belongs needs to be marked, and when marking is required, marks the packet. When the arrived packet belongs to the flow # 21, the packet identifier is marked with a router identifier indicating the control router 30 and a control identifier indicating arrival time designation delivery as control contents.

  If the arriving packet does not require marking, the ingress router 20 forwards the packet to the next router by a normal routing procedure.

  When receiving the marking cancellation request from the communication control device 10, the ingress router 20 deletes the entries for the flow # 21, the flow # 25, and the flow # 54 from the flow management table 25B. At the same time, the packets belonging to these flows are transferred to the next router in the normal procedure without marking. This is because it is not necessary to strictly control the time for simultaneous delivery to the user after a special time such as an auction or a quiz has passed.

  As in the first embodiment shown in FIG. 13, the control router 30 includes a control management unit 111, an execution management unit 112, a flow management unit 113, a packet control unit 114, an input control unit 115, and an output control unit. 116.

  FIG. 19 is a flowchart showing the operation of the control router 30 of the second embodiment. When the input control unit 115 recognizes the arrival of a packet, the execution management unit 112 determines whether the arrival packet has marking information (S601). If there is no marking in the packet (NO in S601), the packet is transferred to the next router as it is by normal routing processing (S609). If the packet is marked (YES in S601), it is determined whether the packet needs to be controlled by the own router (S603).

  If the arriving packet is not subject to control by the own router (NO in S603), the packet is transferred as it is to the next router through normal routing (S609). If it is a control target in its own router (YES in S603), the execution management unit 112 identifies the control content (method) of the arrival packet (S605). The control content can be identified only by determining the control identifier included in the marking information. The packet control unit 114 controls the packet according to the control identifier (S607). In the second embodiment, the control identifier for specifying the time-designated distribution also includes time information, and the control router 30 buffers the packet flows that require time-designated distribution in a batch until a predetermined time is reached. When the time designated by the control identifier comes, the packet is transferred to the next router (S609).

  As described above, in the communication system of the present invention, the flow is controlled in the core part where it is difficult to perform individual control while taking advantage of the features of the QoS control in the unit of flow in the edge part of the network and the scalability by the class control in the core part. Per-packet control is possible.

  In particular, by performing packet control based on the information marked at the ingress router, it is possible to control each flow while minimizing the burden on the router at the core. In addition, it is possible to improve the processing capacity of the router in the core unit.

  In the embodiment, the buffering control triggered by the handover in the mobile communication network and the arrival time designation distribution control have been described as examples. However, the present invention is not limited to these examples. For example, the present invention can be applied to all flow control required in a mobile communication network, such as flow control for multicast communication, packet count control, and retransmission control.

It is a figure for demonstrating the problem of the conventional network system. 1 is a schematic diagram of a system to which the present invention is applied. 1 is a schematic configuration diagram of a communication system 1A according to a first embodiment of the present invention. It is a sequence diagram which shows the flow of a process of the communication system 1A whole. It is a block diagram of the communication control apparatus 10 used with the system of FIG. It is a flowchart which shows the processing operation of the communication control apparatus 10 in the system of FIG. It is a block diagram of the ingress router 20 used with the system of FIG. It is a flowchart which shows the processing operation of the ingress router 20 in the system of FIG. It is a figure which shows the structural example of the flow management table in an ingress router. It is a figure which shows another example of the flow management table in an ingress router. It is a figure which shows another example of the flow management table in an ingress router. It is a figure which shows the process of the ingress router 20 when a marking related request | requirement is received from a communication control apparatus. It is a block diagram of the control router 30 used with the system of FIG. It is a flowchart which shows the processing operation of the control router 30 in the system of FIG. It is a figure which shows the structural example of the control management table used with a control router as needed. It is a figure which shows the process of the control router 30 when a control related request | requirement is received from a communication control apparatus. It is a schematic block diagram of the communication system 1B which concerns on 2nd Embodiment of this invention. It is a flowchart which shows the processing operation of the communication control apparatus 10 in the system of FIG. It is a flowchart which shows operation | movement of the control router used with the communication system of FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1, 1A, 1B Communication system 10 Communication control apparatus 11 Counterpart communication apparatus 12, 12A-12C Mobile terminal 20 Ingress router (edge router)
25, 25A, 25B Ingress router flow management table 30 Control router 31-33 Router 35 Control management table 40A, 40B Base station 50 Mobile communication network 101 Ingress router control management section 102 Ingress router execution management section 103 Ingress router flow management section 104 Ingress Router input control unit 105 Marking control unit 106 Ingress router output control unit 111 Control router control management unit 112 Control router execution management unit 113 Control router flow management unit 114 Packet management unit 115 Control router input control unit 116 Control router output management unit 121 Communication Control management unit 122 of control device Execution management unit 123 of communication control device Flow management unit of communication control device

Claims (11)

A communication control device for managing communication on the network;
An ingress router through which packets transmitted into the network pass;
A control router that performs flow control of the packet inside the network,
The communication control device specifies the ingress router and the control router for each packet flow, and transmits a marking request designating the corresponding control router and the flow control performed by the control router to the ingress router. ,
The ingress router adds marking information including the control router and the flow control to a packet included in the packet flow according to the marking request,
The communication router according to claim 1, wherein the control router controls the packet flow based on marking information of the packet. An execution management unit that detects the necessity of flow control for packets transmitted and received on the network;
For a packet flow that requires flow control, a flow management unit that identifies an ingress router through which the packet flow passes and a control router that performs flow control on the packet flow;
A communication control apparatus comprising: a control management unit that transmits a marking request designating a corresponding control router and flow control performed by the control router to the specified ingress router. The communication control apparatus according to claim 2, wherein the control management unit transmits a control start request to the identified control router. The communication control apparatus according to claim 2, wherein the control management unit transmits a control end request to the specified control router. The communication control apparatus according to claim 2, wherein the control management unit transmits a marking release request to the ingress router when the necessity for the flow control is resolved. A route control device through which a packet transmitted in the network passes,
A control management unit that receives a marking request including a control identifier representing flow control necessary for the packet and a router identifier representing a control router on which the flow control is performed, from a communication control device that manages communication on the network; ,
An execution management unit that identifies a flow to which a packet that has arrived at the routing control device belongs, and determines whether a marking request for the flow has been received;
A routing control device comprising: a marking control unit that adds marking information including the control identifier and a router identifier to the arrived packet when the marking request is received. The path control apparatus according to claim 6 , further comprising a flow management table that stores a control identifier and a router identifier included in the marking request in association with the flow to be subjected to the flow control. The path control device according to claim 6 or 7, wherein the control identifier included in the marking information includes information for distinguishing a flow to be subjected to the flow control. A routing control device for controlling the flow of packets passing through a network,
An execution management unit for determining whether the arrival packet includes marking information for designating flow control in the routing control device;
A route control device comprising: a packet control unit that controls the arrival packet according to the marking information when the arrival packet includes marking information that specifies flow control in the route control device. A control management unit that receives a control start request for a specific packet flow from a communication control device that manages communication on the network;
The path control device according to claim 9 , wherein when the control start request is received, the execution management unit makes preparations necessary for controlling the specific packet flow. Identifying an ingress router through which a packet transmitted into the network passes for each flow of packets transmitted and received on the network, and a control router that performs flow control of the packet within the network;
Sending a marking request designating the corresponding control router and flow control performed by the control router to the ingress router;
In the ingress router, adding marking information specifying the control router and the flow control to a packet passing through the ingress router;
And a step of performing flow control of the packet based on marking information added to the packet in the control router.

Images