JP6542816B2 - Node device and control method thereof - Google Patents

Description

  The present invention relates to a node device and its control method.

  As one of the routing methods, there is a method called AODV (Ad hoc On-Demand Distance Vector) shown in Non-Patent Document 1. This AODV is an ad hoc routing method that is optimal for wireless networks, and as a skeleton of routing by AODV, a node device that requires a route to a destination node device transmits a route search packet by broadcast, and a route The node device having received the search packet forms a route to the transmission source with reference to the information of the route search packet and transfers the route search packet to the surroundings. When a route search packet arrives at a node device having sufficiently new information with respect to the destination node device or the destination node device, the corresponding node device discards the route search packet, and the route to the formed source node device The route reply packet is transmitted by using the above and the route is established by notifying the route to the source node device. When there are a plurality of route candidates, a route with the smallest cost value (number of hops or any value determined for each relay route) of the route is selected.

  When a route search is performed by such AODV, each time a node device receives a route search packet, the cost from the transfer source is added to the cost information in the route search packet, and the route search packet in the received direction After forming a route directed to the node apparatus that has sent the [1], the route search packet is transferred again. When the route search packet reaches a node device that knows sufficiently new information about the route destination or the route destination, the route search packet transfer is aborted, and the route reply information is transmitted to the node device that has sent the route search packet. The node device that has sent the packet can create a route to the destination. Here, the new and old information is managed by a number called a sequence number which each node device has, and is increased each time a route search or answer process regarding itself is executed, and another node device executes the route search process. At the same time, it is recorded with the route information. Thereafter, the other node device can detect the new / old of the information by comparing the sequence number in the newly arrived packet with the already recorded sequence number, and can determine the possibility of updating.

  AODV technology is a routing technology used in multi-hop radio, and each node device broadcasts a route search signal as needed, and an effective route to a node device serving as a target of route search or a target node device The node device that knows the path issues a route reply packet toward the node device that has executed the search, and the node device that has issued the path search packet is notified with the target node device according to the information contained in the path reply packet. Communication can be performed.

IETF RFC 3561 "Ad hoc On-Demand Distance Vector (AODV) Routing" (Experimental, 2003)

  By the way, in AODV, each node device stores a route entry for each destination node, and each node device receives a route entry when receiving a route request (RREQ: Route Request) and a route response (RREP: Route Reply). By storing, a path between node devices is established. Also, the route entry for each destination node stored by each node device is made active if it is frequently used, and becomes inactive when it is not used and a predetermined time (for example, 3 seconds) elapses, and the state is invalid. Is deleted for a predetermined time (for example, 15 seconds).

  For this reason, when receiving a packet from another node device, each node device needs to execute the process of updating the validity period of the corresponding route entry and manage the state of the route entry. When the update process of the expiration date is performed between the reception and the transmission of the packet, the latency of the packet transfer occurs only for the time required for the process. Such problems become more pronounced as the number of node devices configuring the network system increases.

  An object of the present invention is to provide a node device capable of improving the throughput of a network system by reducing the packet transfer latency, and a control method thereof.

In order to solve the above problems, the present invention provides a route entry, which is information indicating a route to another node device, in the node device configuring a network system in which a plurality of node devices are communicably connected by a link. Receiving from a first storage unit storing the expiration date of the route entry , a second storage unit storing a route index having information for uniquely identifying the route entry in use, and another node device Adding means for adding the path index to the second storage means when the path index corresponding to the destination of the received packet is not stored in the second storage means, and the second storage means wherein with respect to the path index of a predetermined number of said stored path index, stored in the first storage unit in a closed- Comprising an updating means for updating the time limit, and deleting means for deleting the path index of updated stored in the second storage means, said updating means, a relay that performs relay processing of the packets If the usage rate of the relay processing unit and / or the control unit is less than a predetermined threshold value with reference to the usage rate of the control unit that controls the processing unit and / or each unit of the node device, the relay processing unit and And / or updating the route entries corresponding to the number of the route indexes according to the usage rate of the control unit .

Further, in the present invention, the updating means is a case where the usage rate of the relay processing unit and / or the control unit is less than a predetermined threshold, and the usage rate of the relay processing unit and / or the control unit is It is characterized in that the path entries corresponding to the path index N1 when low and N2 (<N1) when the utilization rate is high are updated.

Further, according to the present invention, the update unit is configured to use the relay processing unit and / or the control unit even when the usage rate of the relay processing unit and / or the control unit is equal to or more than a predetermined threshold when the route entry is not updated continues for a predetermined time or more. And updating the expiration date of the route entry.

Further, the present invention is characterized in that a usage rate of the relay processing unit and / or the control unit indicates a ratio of time during which processing in a unit time is performed.

Further, in the present invention, the updating unit is configured to execute the packet processing when the number of packets waiting for processing stored in a processing queue storing the packets to be processed by the relay processing unit is less than a predetermined threshold. The path entries of the number corresponding to the number of packets stored in the processing queue among the path indexes stored in the second storage means are updated.

Further, the present invention is characterized in that the second storage means stores the path entry in a FIFO (First In First Out) format.
According to such a configuration, the route entry to be updated can be easily identified.

Further, according to the present invention, there is provided a control method of a node device constituting a network system in which a plurality of node devices are communicably connected by a link, a route entry which is information indicating a route to another node device and the route entry. A second storage unit for storing a route index having information for uniquely identifying the route entry being used, in a first storage step for storing in the first storage unit an expiration date of And adding the path index to the second storage unit if the path index corresponding to the destination of the packet received from the other node device is not stored in the second storage unit. a step, with respect to the path indices of predetermined number of the path index that is stored in the second storage unit, the first Has an updating step which is stored in the storage unit updating the expiration date, and deleting step of deleting the path index updated stored in the second storage unit, wherein the updating step, the When the usage rate of the relay processing unit and / or the control unit is lower than a predetermined threshold with reference to the usage rate of the relay processing unit that executes packet relay processing and / or the control unit that controls each unit of the node device The present invention is characterized in that the route entry corresponding to the number of the route indexes according to the usage rate of the relay processing unit and / or the control unit is updated .

  According to the present invention, it is possible to provide a node device capable of improving the throughput of a network system and its control method by reducing the latency of packet transfer.

It is a figure which shows the structural example of the network system of this invention. It is a figure which shows the example of a detailed structure of the node apparatus shown in FIG. It is a figure for demonstrating the operation | movement of 1st Embodiment. It is a flowchart for demonstrating the operation | movement of 1st Embodiment. It is for demonstrating the operation | movement of 2nd Embodiment. It is a flowchart for demonstrating the operation | movement of 2nd Embodiment. It is a figure for demonstrating the operation | movement of 3rd Embodiment. It is a flowchart for demonstrating 3rd Embodiment. It is a figure for demonstrating the operation | movement of 4th Embodiment. It is a flowchart for demonstrating 4th Embodiment. It is a figure for demonstrating the operation | movement of 5th Embodiment. It is a flowchart for demonstrating 5th Embodiment.

  Next, an embodiment of the present invention will be described.

(A) Description of Configuration of First Embodiment FIG. 1 is a diagram showing an example of a configuration of a network system according to a first embodiment of the present invention. As shown in FIG. 1, the network system according to the first embodiment has node devices 10-1 to 10-11, and node devices 10-1 to 10-11 are connected by links. In the example of FIG. 1, the node devices 10-1 to 10-11 are configured not as mobiles but as fixed devices. The node devices 10-1 to 10-11 execute routing processing based on AODV.

  FIG. 2 shows a detailed configuration example of the node device. Since the node devices 10-1 to 10-11 have the same configuration, hereinafter, these will be described as the node device 10. As illustrated in FIG. 2, the node device 10 includes a packet relay processing unit 11, a control unit 12, a storage unit 13, reception units 14-1 to 14-n, and transmission units 15-1 to 15-n. ing. In the example of FIG. 1, for example, two links are connected to the node device 10-3, and therefore, among the receiving units 14-1 to 14-n and the transmitting units 15-1 to 15-n, the link is selected. Two are connected.

  Here, under the control of the control unit 12, the packet relay processing unit 11 transmits corresponding packets received by the receiving units 14-1 to 14-n according to the information stored in the header thereof. Send from the units 15-1 to 15-n. The control unit 12 rewrites the header of the received packet according to the route table 13a stored in the storage unit 13, relays the packet via the packet relay processing unit 11, and transmits a route request (RREQ) packet and a route. Perform processing on reply (RREP) packets.

  The storage unit 13 is constituted by a semiconductor memory, and is related to a route table 13a, which is information for transferring packets, a management table 13b for managing route entries stored in the route table 13a, and route search described later. It stores programs and data for executing processing. The receiving units 14-1 to 14-n receive packets via links. Also, the transmitting units 15-1 to 15-n transmit packets via links. The receiving unit and the transmitting unit in which the same number is given after the hyphen are connected to the same link. For example, the receiving unit 14-1 and the transmitting unit 15-1 are connected to the same link.

(B) Description of the Operation of the First Embodiment Next, the operation of the first embodiment of the present invention will be described. In the following, in the network system shown in FIG. 1, it is assumed that the node device 10-2 makes a route request to search for a route to the node device 10-7. In that case, the source node device 10-2 floods the RREQ with the node device 10-7 as the destination address. The node devices 10-3 to 10-6, 10-8 and 10-9 receiving the RREQ also flood the same RREQ.

  At this time, when transferring the RREQ, each node device associates the source address (the address of the node device 10-2) contained in the packet with the address of the immediately preceding node device as a path entry, It is stored in the 13 route tables 13a. For example, in the present example, since the node device 10-4 receives the RREQ from the node device 10-3, the address of the node device 10-2 as the transmission source address and the address of the node device 10-3 can be obtained. It associates and stores it as a route entry in the route table 13a of the storage unit 13. The source address of the route entry registered in this way is used as a destination address, and the address of the immediately preceding node device is used as the next transfer destination (next hop) of the packet.

  The RREQ transmitted from the node device 10-2 arrives at the node device 10-7 via clockwise and counterclockwise paths. The node device 10-7 has a cost value (for example, a node device that passes through the counterclockwise path (the node device 10-2, the node device 10-8, the node device 10-9, and the node device 10-7). And, since the number of links) is small, an RREP having the node device 10-2 as a transmission destination address is returned in unicast via the node device 10-9. The RREP is delivered to the source node device 10-2 with reference to the route entry generated upon transfer of the RREQ.

  At this time, when transferring the RREP, each node device associates the source address (the address of the node device 10-7) contained in the packet with the address of the immediately preceding node device as a path entry, It is stored in the 13 route tables 13a. For example, since the node device 10-3 receives the RREP from the node device 10-4, the node device 10-3 stores the address of the source node address of the node device 10-7 and the address of the node device 10-4 in association with each other. The path table 13a of the unit 13 is stored as a path entry. The source address of the path entry registered in this way is used as a destination address, and the address of the immediately preceding node device is used as the next transfer destination (next hop) of the packet. Note that, as a precursor list (PL: Precursor List) for transmitting a route error (RERR: Route Error), the immediately preceding node device may be registered together with the route entry. For example, in the case of the node device 10-3, for the route entry whose transmission destination is the node device 10-7, it is possible to register the node device 10-2 that corresponds immediately before as a PL. Similarly, in the case of the node device 10-3, for the route entry whose transmission destination is the node device 10-2, it is possible to register the node device 10-4, which corresponds immediately before, as the PL.

  By the above operation, a path is established between the node device 10-2 and the node device 10-7. Thus, when a path is established between the node device 10-2 and the node device 10-7, the node device 10-2 starts transmission of data packets to the node device 10-7.

  A route entry is stored in the route table 13a of the storage unit 13 in response to the transfer of RREQ and RREP. In the prior art, such a route entry is in an "invalid" state when a predetermined time (for example, 3 seconds) elapses without the corresponding route entry being used, and an invalid state is a predetermined time (for example, 15 seconds) ) Is deleted from the route table 13a. Also, the route entry has its expiration date reset (e.g., set to 0) when the route entry is used, and restarts counting of 3 seconds and 15 seconds described above. For this reason, in the prior art, since it is necessary to update (reset) the expiration date of the corresponding route entry every time the route is used, there is a problem that, for example, processing load is applied to the control unit 12.

  On the other hand, in the first embodiment of the present invention, each node device has the management table 13b for managing the route entry stored in the route table 13a, and based on the management table 13b, the route table Execute the update process of the expiration date of the route entry of 13a. FIG. 3 is a diagram for explaining the operation of the management table 13b. As shown in FIG. 3A, the management table 13b stores a route index corresponding to the route entry stored in the route table 13a. For example, the route index A shown in FIG. 3A is an index corresponding to the route entry A (not shown) stored in the route table 13a.

  When a packet addressed to another device is received, the control unit 12 searches whether or not the corresponding route entry is stored in the route table 13a. If the packet is stored, the packet is transferred using the route entry. Do. In the present embodiment, at this time, the update processing of the validity period of the corresponding route entry is not executed. On the other hand, when no route entry is stored in the route table 13a, RREQ is executed to perform a route search, and the route entry obtained as a result is stored in the route table 13a.

  Next, the control unit 12 determines whether or not the route index corresponding to the used (or newly generated) route entry is stored in the management table 13b, and the corresponding route index is not stored. Creates and stores a path index. For example, when the route entry E is used, if the route index E corresponding to the route entry E is not stored in the management table 13b, as shown in FIG. 3A, the route index E is not stored in the management table 13b. Register at the end. Here, the route index is information for identifying a route entry stored in the route table 13a (for example, ID (Identification) information for uniquely identifying a route entry and a route entry are stored). Address information indicating an address). For example, taking the route index E as an example, information indicating the address of the route entry E stored in the route table 13a is stored in the route index E.

  The management table 13b is updated at predetermined time intervals. That is, when a predetermined time has elapsed since execution of the update processing, the control unit 12 updates the expiration date of the route entry corresponding to all the route indexes stored in the management table 13b, and updates the expiration date. The route index completed by the above is deleted from the management table 13b. For example, in the example of FIG. 3B, the expiration date of the route entry A stored in the route table 13a is updated with reference to the ID or the address stored in the route index A stored at the left end. As a method of updating, for example, there is a method of extending the expiration date. That is, the route entry is deleted from the route table 13a as being expired when a predetermined time (for example, 15 seconds) elapses because the unused state continues, but the route index targeted for the update process is For the corresponding route entry, the expiration date is extended, and a predetermined time is counted from 0 seconds. In addition, it may be made to set to predetermined time (for example, 1 second) instead of 0 second.

  The control unit 12 collectively executes the updating process of the validity period of the route entry corresponding to all the route indexes stored in the management table 13b (see FIG. 3B). As a result, when the update processing of the validity period of the route entry corresponding to all the route indexes is completed, the control unit 12 deletes all the route indexes stored in the management table 13b. As a result, no route index is stored in the management table 13b (see FIG. 3C).

  The time when the management table 13b is updated needs to be set shorter than the expiration date of the route entry stored in the route table 13a. For example, if the route entry expiration date is 15 seconds, it should be set to less than 15 seconds. This is because when updating the route entry based on the route index, the route entry is deleted if the route entry expiration date is exceeded.

  Next, the detailed operation of the first embodiment will be described with reference to FIG. FIG. 4 is a flowchart for explaining an example of processing executed in each node device shown in FIG. When the flowchart shown in FIG. 4 is started, the following steps are performed.

  In step S10, the control unit 12 resets a built-in timer (not shown). As a result, the count value of the timer is set to "0", and the timer starts counting the elapsed time.

  In step S11, the control unit 12 determines whether or not a packet whose transmission destination is another node device has been received, and if it is determined that it has been received (step S11: Y), the process proceeds to step S12 In the case of (step S11: N), the process proceeds to step S15.

  In step S12, the control unit 12 executes analysis processing of the packet received in step S11. More specifically, the control unit 12 analyzes the destination of the packet by referring to the header information of the packet. When the analysis process is completed, the control unit 12 searches the route table 13a, determines whether or not the corresponding route entry exists, and executes the route search process if the route entry does not exist. Also, if there is a corresponding route entry, packet transfer processing is executed with reference to the corresponding route entry. At this time, the process of updating the route entry expiration date is not executed.

  In step S13, the control unit 12 searches the management table 13b of the storage unit 13 based on the information obtained as a result of the analysis in step S12, and determines whether or not there is a corresponding route index. If it is determined that there is a corresponding route index (step S13: Y), the process proceeds to step S15. Otherwise (step S13: N), the process proceeds to step S14.

  In step S14, the control unit 12 registers as a new route index because the corresponding route index does not exist in step S13. As the route index, the address of the route table 13a in which the corresponding route entry is stored, or ID information capable of uniquely identifying the route entry can be used. Of course, other information may be used.

  In step S15, the control unit 12 refers to the count value of the timer reset in step S10 (or step S16) and determines whether a predetermined time (for example, 14 seconds) has elapsed since the reset. If it is determined that the predetermined time has elapsed (step S15: Y), the process proceeds to step S16. Otherwise (step S15: N), the process proceeds to step S19.

  In step S16, the control unit 12 resets the timer. As a result, the count value of the timer is set to "0", and the timer starts counting the elapsed time.

  In step S17, the control unit 12 executes a process of updating the term of validity of the route entry corresponding to the route index stored in the management table 13b. For example, in the example of FIG. 3B, the expiration date of each of the route entries A to E corresponding to the route indexes A to E is updated. As a specific example of the renewal of the expiration date, for example, there is a method of setting 0. It may be set to a value other than 0.

  In step S18, the control unit 12 deletes all route indexes stored in the management table 13b of the storage unit 13. As a result, the management table 13b is empty as shown in FIG. 3 (C).

  In step S19, the control unit 12 determines whether to end the process, and when it is determined not to end the process (step S19: N), the process returns to step S11 and the same process as described above is repeated. If not (step S19: Y), the process ends.

  According to the above processing, the above-described operation can be realized.

  As described above, according to the first embodiment of the present invention, the management table 13b is provided to store the route index corresponding to the route entry in use, and a predetermined time has elapsed since the previous update. In this case, since the validity period of the route entry corresponding to the route index stored in the management table 13b is collectively updated, it is compared to the case where the validity period of the route entry is updated each time a packet is received. The processing of the control unit 12 can be reduced. This makes it possible to reduce the packet transfer latency and improve the throughput of the network system.

(C) Description of Configuration of Second Embodiment Next, a second embodiment of the present invention will be described. In the second embodiment, as compared with the first embodiment, the management table 13 b is a FIFO (First In First Out) type table. Also, the processing executed by the control unit 12 is different. Since the configuration other than these is the same as that of the first embodiment, the detailed description will be omitted.

(D) Description of Operation of Second Embodiment FIG. 5 is a diagram for explaining an operation of the second embodiment. As shown in FIG. 5A, when each node device receives a packet whose destination is another node device, each node device confirms that the same route index does not exist, and then manages the route index management table. Add to the end of 13b. In the example of FIG. 5A, a route index E (for example, a route index corresponding to a route entry whose destination is the node device 10-7 in FIG. 1) is newly added. Since the path index E does not exist in the management table 13b, the path index E is added to the end of the FIFO-type management table 13b.

  If a predetermined time has elapsed after updating the management table 13 b (for example, if 14 seconds have elapsed), the control unit 12 may, for example, select one of the route indexes A to E shown in FIG. The validity period of the route entry corresponding to a predetermined number of predetermined route indexes (three route indexes A, C, E in the example of FIG. 5) is updated, and the route indexes A, C, E are collectively delete. When the path index is deleted, the FIFO management table 13b moves the path index to the right in the figure, so as shown in FIG. 5 (C), the path index B, Only E exists.

  Here, for example, a value set by default can be used as the number of route index deletions. Besides this, for example, the administrator may be able to set a desired value, or the optimum value may be automatically set according to the delay state or latency of the network.

  When the corresponding route entry does not exist, the route update process (the process by the above-described RREQ and RREP) is executed, the route entry is generated again, and stored in the route table 13a. An index is generated and stored in the management table 13b.

  Next, the detailed operation of the second embodiment will be described with reference to FIG. FIG. 6 is a flowchart for explaining an example of processing executed in each node device shown in FIG. In addition, in FIG. 6, since the same code | symbol is attached | subjected to the part corresponding to FIG. 4, the description is abbreviate | omitted. In FIG. 6, step S17 and step S18 are replaced with step S30 and step S31 as compared with FIG. The processes other than these are the same as in the case of FIG. Hereinafter, the operation will be described focusing on step S30 and step S31.

  The control unit 12 determines in step S15 that the predetermined time has elapsed, and resets the timer in step S16, and then in step S30, the predetermined number of path indexes stored in the management table 13b of the storage unit 13 Update the expiration date of the corresponding route entry. For example, the control unit 12 is a process of updating the expiration date of the route entries A, C, E corresponding to the three route indexes (route indexes A, C, E in the example of FIG. 5) stored in the management table 13b. Run. As a result, the expiration date of the route entries A, C, and E stored in the route table 13a is updated.

  In step S31, the control unit 12 executes a process of deleting the route index corresponding to the route entry whose validity period has been updated in step S30 from the management table 13b. As a result, in the example of FIG. 5, the route indexes B and D which are not targets of renewal of the expiration date move to the right side of FIG. 5, and are in the state of FIG.

  As described above, according to the second embodiment of the present invention, only the unstored route index is added to the management table 13b, and the route stored in the management table 13b when a predetermined time has elapsed. By updating the expiration date of the route entry corresponding to a part of the index and deleting the route index corresponding to the route entry targeted for updating, there is no need to update the expiration date each time a packet is received. . Therefore, by reducing the load on the control unit 12, it is possible to reduce the packet transfer latency and to improve the throughput of the network system. Further, as in the first embodiment, when all the route entries are updated collectively, the load on the control unit 12 increases at the timing of the update, but in the second embodiment, since the load is updated in a divided manner, Load can be distributed.

(E) Description of Configuration of Third Embodiment Next, a third embodiment of the present invention will be described. In the third embodiment, compared to the first embodiment, the management table 13 b is a FIFO format table. Also, the processing executed by the control unit 12 is different. Since the configuration other than these is the same as that of the first embodiment, the detailed description will be omitted.

(F) Description of Operation of Third Embodiment FIG. 7 is a view for explaining an operation of the third embodiment. When each node device receives a packet whose destination is another node device as shown in FIG. 7A, after confirming that the same route index does not exist, the route index management table 13b is used. Add to the end of In the example of FIG. 7A, the route index E (for example, the route to the node device 10-7 of FIG. 1 as a transmission destination) is newly added. However, the route index E is added to the management table 13b. Since it does not exist, the route index E is added to the end of the FIFO management table 13b.

  In the third embodiment, when the route entry expiration date is updated based on the route index, the timing and number of update are determined according to the usage rate of the control unit 12 and the usage rate of the packet relay processing unit 11. Ru. More specifically, when the usage rate U1 of the control unit 12 and the usage rate U2 of the packet relay processing unit 11 are obtained, and the usage rates thereof become less than predetermined threshold values Th1 and Th2, respectively, U1 and U2 A predetermined number of route entries are updated according to the value of. If the state in which the term of validity is not updated continues for a predetermined time or more, processing for updating the term of validity is executed regardless of the values of U1 and U2 in order to prevent the path entry in use from being deleted. You may do so.

  Next, the detailed operation of the third embodiment will be described with reference to FIG. FIG. 8 is a flowchart for explaining an example of processing executed in each node device shown in FIG. In addition, in FIG. 8, since the same code | symbol is attached | subjected to the part corresponding to FIG. 4, the description is abbreviate | omitted. In FIG. 8, compared with FIG. 4, steps S <b> 15 to S <b> 18 are replaced with steps S <b> 50 to S <b> 54. The processes other than these are the same as in the case of FIG. Hereinafter, the operation will be described focusing on steps S50 to S54.

  In step S50, the control unit 12 acquires its own usage rate U1. Here, the "usage rate" indicates the percentage of time during which a certain process is being performed within a predetermined unit time (for example, 1 second), and if the utilization rate is 0%, nothing is executed and the process waits In the case of 100%, it indicates that some processing is being continued.

  In step S51, the control unit 12 acquires the usage rate U2 of the packet relay processing unit 11. The usage rate indicates the proportion of time during which the packet relay processing unit 11 is executing some processing within a predetermined unit time, as in the above-described case.

  In step S52, the control unit 12 determines whether U1 is less than a predetermined threshold Th1 and whether U2 is less than a predetermined threshold Th2, and if U1 <Th1 and U2 <Th2 are satisfied (step S52) If Y (Y), then the process proceeds to step S53; otherwise (N in step S52), the process proceeds to step S19. If the state in which the validity period is not updated continues for a predetermined time or more, the process proceeds to step S53 to update the validity period regardless of the values of U1 and U2 in order to prevent the route entry in use from being deleted. Processing may be performed.

  In step S53, the control unit 12 updates the validity period of the route entry corresponding to the number of route indexes according to U1 and U2. For example, if U1 and U2 have values close to 0%, the expiration date of N1 path entries is updated, and if U1 and U2 have values close to threshold values Th1 and Th2, N2 (N2 <N1) The expiration date of the route entry can be updated. The specific values of N1 and N2 may be set in advance (set statically), or may be set (set dynamically) according to the state of the node device or the network system. Good.

  In step S54, the control unit 12 executes a process of deleting the route index corresponding to the route entry for which the update process of the expiration date is performed in step S53. For example, in the example of FIG. 7, route indexes A and B are deleted from the management table 13b.

  As described above, according to the third embodiment of the present invention, a part of the route entry is collectively updated according to the usage rates of the control unit 12 and the packet relay processing unit 11. There is no need to update the route entry expiration date each time a packet is received. Therefore, by reducing the load on the control unit 12, it is possible to reduce the packet transfer latency and to improve the throughput of the network system. In addition, since the update process is performed when the usage rates of the control unit 12 and the packet relay processing unit 11 fall below the predetermined thresholds Th1 and Th2, the update process may be performed by selecting a timing at which the usage rate is low. it can. Further, since the number of updates is made variable according to the usage rates of the control unit 12 and the packet relay processing unit 11, when the usage rate is low, the number of path entries is relatively high. The load of the control unit 12 or the like can be distributed by updating a small number of path entries.

  In the above description, the usage rates U1 and U2 of both the control unit 12 and the packet relay processing unit 11 are referred to, but the route entry is updated with reference to the usage rates of only one of them. It may be determined whether or not to carry out and / or the number of route entries to be updated. In step S53, the number of path entries according to U1 and U2 is updated, but for example, U1 and U2 are acquired as needed, and update processing is performed until U1 and U2 exceed predetermined threshold values Th1 and Th2. You may make it continue. Of course, at that time, the valid time of the route entry in use may also be referred to.

(G) Description of Configuration of Fourth Embodiment Next, a fourth embodiment of the present invention will be described. In the fourth embodiment, compared with the first embodiment, the management table 13 b is a FIFO type table. Also, the processing executed by the control unit 12 is different. Since the configuration other than these is the same as that of the first embodiment, the detailed description will be omitted.

(H) Description of the Operation of the Fourth Embodiment Next, the operation of the fourth embodiment of the present invention will be described. FIG. 9 is a diagram for explaining the operation of the fourth embodiment. When each node device receives a packet whose destination is another node device as shown in FIG. 9A, after confirming that the same route index does not exist, the route index is managed in the management table 13b. Add to the end of

  In the fourth embodiment, when the route index is deleted (and the corresponding route entry is updated), as shown in FIG. 9B, the timing and number of deletions are the processing queues of the packet relay processing unit 11. It is determined according to the number of packets (N) in More specifically, the packet relay processing unit 11 stores the received packets in the processing queue, and executes processing (analysis processing, transfer processing, and the like) in order. If there are a large number of waiting packets in the processing queue, this indicates that the load status is high. In this case, route entry update processing is put on hold, or a small number of route entries are updated. . Further, when the number of packets in the standby state existing in the processing queue is small, it indicates that the load state is low, and in that case, update processing of many route entries is executed. In the example of FIG. 9C, the route indexes A, C, and E are deleted (at this time, the route entries A, C, and E are updated). As described above, by executing the route entry update process according to the number of packets stored in the process queue, the latency can be reduced and the throughput can be improved.

  Next, the detailed operation of the fourth embodiment will be described with reference to FIG. FIG. 10 is a flowchart for explaining an example of processing executed in each node device shown in FIG. In FIG. 10, the parts corresponding to those in FIG. 4 are given the same reference numerals, and the description thereof will be omitted. In FIG. 10, compared with FIG. 4, step S15 to step S18 are replaced with step S70 to step S73. The processes other than these are the same as in the case of FIG. Hereinafter, the operation will be described focusing on steps S70 to S73.

  In step S70, the control unit 12 obtains the number N of packets in the standby state stored in the processing queue (not shown) of the packet relay processing unit 11.

  In step S71, the control unit 12 determines whether the number N of packets in the processing queue acquired in step S70 is less than a predetermined threshold Th, and if N <Th is satisfied (step S71: Y) The process proceeds to step S72, and otherwise (step S71: N), the process proceeds to step S19. The specific value of N may be set in advance (set statically) or may be set dynamically according to the operating state of the network system or the node device. If the state in which the term of validity is not updated continues for a predetermined time or more, the process proceeds to step S72 to update the term of validity regardless of the value of N in order to prevent the route entry in use from being deleted. May be performed.

  In step S72, the control unit 12 executes the process of updating the validity period of the route entry corresponding to the number of route indexes according to the value of N acquired in step S70. For example, when the number N of packets in the processing queue is small, more path entries are updated, and when the number N is large, the smaller path entries are updated. The number of route entries to be specifically updated may be set in advance (set statically) or may be set dynamically according to the operating state of the network system or the node device.

  As described above, according to the fourth embodiment of the present invention, a part of the route entry is collectively updated according to the number of packets in the waiting state in the processing queue. There is no need to update the route entry expiration date each time it is received. Therefore, by reducing the load on the control unit 12, it is possible to reduce the packet transfer latency and to improve the throughput of the network system. Further, since the update process is performed when the number N of packets in the processing queue falls below the predetermined threshold value Th, the update process can be performed by selecting a timing at which the load is low. Also, since the number of updates is variable according to the number N of packets in the processing queue, many path entries are low when the load is low, and a small number of path entries are high when the load is relatively high. The load on the control unit 12 can be further reduced by updating.

  In the above description, although the number of route entries according to N is updated in step S72, for example, N is acquired as needed, and the update process is continued until N exceeds a predetermined threshold Th. You may

(I) Description of Configuration of Fifth Embodiment Next, a fifth embodiment of the present invention will be described. In the fifth embodiment, as compared with the first embodiment, the management table 13 b is a FIFO format table. Also, the processing executed by the control unit 12 is different. Since the configuration other than these is the same as that of the first embodiment, the detailed description will be omitted.

(J) Description of the Operation of the Fifth Embodiment Next, the operation of the fifth embodiment of the present invention will be described. FIG. 11 is a diagram for explaining the operation of the fifth embodiment. When each node device receives a packet whose destination is another node device as shown in FIG. 11A, after confirming that the same route index does not exist, the route index is managed in the management table 13b. Add to the end of

  In the fifth embodiment, when the route entry is updated, as shown in FIG. 11B, the timing of the update is determined by a timer, and the number of updates is determined by the expiration date of the route entry. More specifically, if a predetermined time has elapsed after updating the management table 13b (for example, if 14 seconds have elapsed), the expiration date of the route indexes A to E shown in FIG. Updates the expiration date of the route entry below a predetermined value (e.g., T). As an example, when the expiration date is set to 15 seconds and the route entry whose expiration date is less than 10 seconds is to be updated, the control unit 12 updates the route entry whose expiration date is less than 10 seconds. In the example of FIG. 11B, three route entries A, C, and E whose expiration date is less than T are collectively updated. When the route entry is updated, the corresponding route index is deleted from the management table 13b. When the path index is deleted, the FIFO management table 13b moves the path index to the right in the figure, so as shown in FIG. 11C, only the path entries B and E are stored in the management table 13b. Is present. As described above, by periodically executing the route entry update processing by the timer, the latency can be reduced and the throughput can be improved. In addition, since the route entry whose validity period is less than T is updated when updating is performed, the load of the control unit 12 is concentrated at a predetermined timing by distributing and executing the update processing temporally. Can be prevented.

  Next, the detailed operation of the fifth embodiment will be described with reference to FIG. FIG. 12 is a flowchart for explaining an example of processing executed in each node device shown in FIG. In FIG. 12, the parts corresponding to those in FIG. 4 are given the same reference numerals and thus the description thereof is omitted. In FIG. 12, as compared with FIG. 4, step S17 and step S18 are replaced with step S90 and step S91. The processes other than these are the same as in the case of FIG. In the following, the operation will be described focusing on step S90 and step S91.

  After determining that the predetermined time has elapsed in step S15 and resetting the timer in step S16, the control unit 12 determines, in step S90, a route entry corresponding to the route index stored in the management table 13b of the storage unit 13. And renew the expiration date of the route entry whose expiration date is less than T. For example, among the route entries corresponding to the route index stored in the management table 13b, the control unit 12 updates the valid period of the route entry whose valid period is less than a predetermined time (for example, 10 seconds). As an example, when the expiration date is set to 15 seconds and a route entry whose expiration date is less than 10 seconds is to be updated, the control unit 12 updates path entries whose expiration date is less than 10 seconds.

  In step S91, the control unit 12 executes a process of deleting from the management table 13b the route index corresponding to the route entry for which the update process of the expiration date has been executed in step S90. For example, in the example of FIG. 11, the route indexes A, C, and E corresponding to the route entries A, C, and E whose validity dates have been updated are deleted from the management table 13b.

  As described above, according to the fifth embodiment of the present invention, part of the route entry is collectively updated when a predetermined time has elapsed, so that the expiration date is received each time a packet is received. There is no need to update Therefore, by reducing the load on the control unit 12, it is possible to reduce the packet transfer latency and to improve the throughput of the network system. Further, as in the first embodiment, when all the route entries are updated collectively, the load on the control unit 12 increases at the timing of the update, but in the fifth embodiment, the load is divided and updated. Load can be distributed.

(K) Description of Modified Embodiments It goes without saying that each of the above-described embodiments is an example, and the present invention is not limited only to the case described above. For example, in each of the above embodiments, as illustrated in FIG. 1, the network system to which 11 of the node devices 10-1 to 10-11 are connected is described as an example, but 10 or less or 12 or more The present invention may be applied to a network system in which node devices are connected.

  In the first, second, and fifth embodiments, the predetermined time in step S15 is fixed. However, the predetermined time may be set to any value. Alternatively, the time in step S15 may be automatically changed according to the number of route indexes. For example, the predetermined time may be adjusted according to the status of communication and the like. According to such a configuration, it is possible to update the route entry at an optimal time according to the communication status.

  In each of the above embodiments, the node devices 10-1 to 10-11 are configured to be fixed not to move, but of course, these may be configured as mobiles. Of course, not all may be mobiles, but some may be mobiles.

  In each of the above embodiments, the destination node device transmits the route reply packet, but in addition to this, the node device having information on the route to the destination node device is The route reply packet may be transmitted. In that case, as in the case of each of the above-described embodiments, the route reply packet may be transmitted to all the node devices connected to this node device.

  In the above second to fifth embodiments, the management table 13b is described as a FIFO type memory, but of course, a memory other than the FIFO type may be used. For example, by using an ordinary memory and managing an address for storing a route entry, the same function as the FIFO may be provided. Also in the first embodiment, the management table 13b may be a FIFO memory.

  In each of the above embodiments, although the case of using the transmission destination address, the next hop, and the precursor as the route entry has been described as an example, other than these, for example, the transmission destination sequence number, transmission The valid flag of the destination sequence number, other flags, the network interface, the number of hops, and the expiration date may be registered.

10-1 to 10-11 node 11 packet relay processing unit 12 control unit (addition means, update means, deletion means)
13 storage unit (first storage unit, second storage unit)
13a Route Table 14-1 to 14-n Receiver 15-1 to 15-n Transmitter

Claims (7)

In the node device constituting a network system in which a plurality of node devices are communicably connected by a link,
A route entry, which is information indicating a route to another node device, and a first storage unit for storing an expiration date of the route entry;
Second storage means for storing a route index having information for uniquely identifying the route entry being used;
Adding means for adding the path index to the second storage means when the path index corresponding to the transmission destination of the packet received from another node device is not stored in the second storage means;
Updating means for updating the expiration date stored in the first storage means with respect to a predetermined number of the path indexes among the path indexes stored in the second storage means;
And deletion means for deleting the updated path index stored in the second storage means ,
The updating unit refers to the usage rate of the relay processing unit that executes the relay processing of the packet and / or the control unit that controls each unit of the node device, and the usage rate of the relay processing unit and / or the control unit is predetermined. And updating the route entry corresponding to the number of route indexes according to the usage rate of the relay processing unit and / or the control unit.
A node device characterized by The updating unit is used when the usage rate of the relay processing unit and / or the control unit is less than a predetermined threshold, and N1 is used when the usage rate of the relay processing unit and / or the control unit is low. The node apparatus according to claim 1, characterized in that when the rate is high, the path entry corresponding to the N2 (<N1) path indices is updated. The update unit is configured to update the route entry even when the utilization rate of the relay processing unit and / or the control unit is equal to or greater than a predetermined threshold when the state in which the route entry is not updated continues for a predetermined time or more. The node device according to claim 1, wherein the expiration date is updated. The node device according to any one of claims 1 to 3, wherein a usage rate of the relay processing unit and / or the control unit indicates a ratio of time during which processing is performed in unit time. . The updating unit is configured to use the second storage unit when the number of the packets waiting for processing stored in the processing queue storing the packets to be processed by the relay processing unit is less than a predetermined threshold. 5. The method according to any one of claims 1 to 4, characterized in that the number of path entries corresponding to the number of packets stored in the processing queue among the path indexes stored is updated. Node device as described. The node apparatus according to any one of claims 1 to 5, wherein the second storage means stores the path entry in a FIFO (First In First Out) format. In the control method of the node device configuring a network system in which a plurality of node devices are communicably connected by a link,
A path entry which is information indicating a path to another node device, and a first storing step for storing an expiration date of the path entry in a first storage unit;
Storing in the second storage unit a route index having information for uniquely identifying the route entry being used;
Adding the path index to the second storage unit if the path index corresponding to the transmission destination of the packet received from another node device is not stored in the second storage unit;
An updating step of updating the expiration date stored in the first storage unit with respect to a predetermined number of the path indexes among the route indexes stored in the second storage unit;
Deleting the updated path index stored in the second storage unit;
The updating step refers to the usage rate of the relay processing unit that executes the relay processing of the packet and / or the control unit that controls each unit of the node device, and the usage rate of the relay processing unit and / or the control unit is predetermined. And updating the route entry corresponding to the number of route indexes according to the usage rate of the relay processing unit and / or the control unit.
And controlling a node device.

Images