JP2014003683A - Packet relay system and radio node - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a packet relay system and radio node capable of quickly relaying a packet to a destination node even when link instantaneous interruption occurs in a path to a sink node.SOLUTION: A radio node acquires an address of one of adjacent nodes adjacent to the radio node and addresses of 2 hop adjacent nodes adjacent to the one adjacent node, from the one adjacent node. When it is determined that a data packet has not been relayed to one of the adjacent nodes, the data packet is relayed to destinations including an address of the one adjacent node and one of addresses of adjacent nodes corresponding to one of addresses of the 2 hop adjacent nodes.

Description

  The present invention relates to a packet relay system including a plurality of nodes that relay data packets to each other.

  There is known a multi-hop wireless network in which a network is configured by a plurality of wireless nodes having a packet relay function and packets are mutually relayed between adjacent wireless nodes. By sequentially relaying packets between wireless nodes, packets can be transferred to wireless nodes where radio waves do not reach directly.

  In order to relay a packet to a desired wireless node, a routing protocol for determining a wireless node to be transferred next from a destination address is required. For example, in the ZigBee (registered trademark) alliance, a cluster tree routing method that eliminates the need for a routing table by devising an address assignment method, an AODV (Ad Hoc On-Demand Distance Vector) method that can configure a mesh topology, multiple A routing protocol based on the many-to-one method that can efficiently collect data from any node is standardized. In particular, the many-to-one method is a method that considers 1-to-n communication between a sink node and a plurality of nodes, and is expected to be applied to applications such as sensing environmental information and automatic meter reading of meters. Yes.

  FIG. 22 is a schematic diagram showing a mode of packet relay between nodes at the time of routing in the many-to-one method. FIG. 22A is a schematic diagram illustrating an aspect of RREQ (Route Request) packet relay when each of the nodes b to j sets an uplink route for relaying a packet to the node a which is a sink node. . The node a transmits an RREQ packet using the transmission destination address as a broadcast address. The node that has received the RREQ packet relays the RREQ packet to another node. The RREQ packet includes a path cost that represents a logical distance between any two nodes on the network. When relaying the RREQ packet, the link cost for the node adjacent to the RREQ packet is added to the path cost and transmitted. The link cost represents a logical distance between nodes on the network, and is a cost that is a basis for calculating a path cost. The link cost is calculated from the reception strength between nodes and the packet error rate. By repeating the relay of the RREQ packet in this way, each of the nodes b to j can set an uplink path for transferring the packet to the sink node a. That is, each of the nodes b to j can relay the packet to the sink node under the optimum conditions by selecting, for example, the node corresponding to the lowest path cost as the packet relay destination node.

  FIG. 22B is a schematic diagram showing a mode of RREC (Route Record) packet relay when setting a downlink route from the node a which is the sink node to the node h. The node h sets the destination as the node a and transmits an RREC packet including its own address. The node that has received the RREC packet transfers the RREC packet including its own address and the number of relays to the transfer destination node. That is, a packet with addresses added in the order of d, c, b, a from node h is transferred to node a. The same processing is performed when setting a downlink route from node a to another node.

  FIG.22 (c) is a schematic diagram showing the mode of packet relay from the node a to the node h after setting a downlink route. The node a transmits a data packet with the address in the reverse order to the address addition order at the time of setting the downlink route described in the packet header. The node that received the data packet deletes its own address described in the packet header and transfers it to the transfer destination node. In this way, data packets can be relayed in the order of nodes a to b, c, d, and h. These routings are called source routing because the source node (source node) is path control that uniquely determines a node that relays a packet.

ZigBeeSpecification Revision17 (ZigBee Document 053474r17

  However, in the conventional many-to-one routing, the link cost calculation method is not clear. In Non-Patent Document 1, the link cost is calculated from factors such as LQI (Link Quality Indicator). However, since the LQI when receiving the RREQ command varies, it is difficult to calculate an accurate link cost. There was a problem of being. In particular, when nodes are installed in a wide area such as outdoors, when a shielding object such as a person or a vehicle passes between the nodes, a momentary disconnection of the link between the nodes occurs, and the link state changes. Therefore, there is a demand for a system that can calculate an appropriate link cost even when the link status fluctuates.

  Further, in the routing in the conventional many-to-one method, in order to establish an upstream path to the sink node, it is necessary to periodically transmit the RREQ packet as shown in FIG. 22 (a). Since the RREQ packet is retransmitted once at each node and delivered to the entire network, there is a problem that the amount of communication information in the network increases and the use efficiency of the radio band deteriorates as the network scale increases. Therefore, there is a demand for a system that can establish an upstream path to a sink node without increasing the amount of communication information in the network.

  Further, in the routing in the conventional many-to-one system, in order to establish a downstream path from the sink node, each node transmits an RREC command to the sink node by unicast transmission as shown in FIG. 22 (b). When the transmission timings at each node overlap, there is a problem that the traffic around the sink node is congested. Therefore, there is a demand for a system that can establish a downstream path from the sink node without causing congestion in the traffic.

  Also, in the conventional many-to-one routing, if the packet is not correctly transferred to the destination node due to changes in the connection state of the node or instantaneous link breaks, the downstream path from the sink node is re-established. Therefore, the RREC command is transmitted from the destination node to the sink node. When such a process is performed, there is a problem that the downlink communication as shown in FIG. 22C cannot be performed until the downlink route is re-established. Therefore, there is a demand for a system that can quickly transmit a packet to a destination node even when packet relay fails in the middle of a downstream path from the sink node.

  An object of the present invention is to provide a packet relay system that can quickly transmit a packet to a destination node even when the packet is not correctly transferred to the destination node in a downstream path from the sink node.

  A packet relay system according to the present invention is a packet relay system that includes a plurality of nodes that relay data packets to each other, and each of the nodes is associated with one of the adjacent nodes adjacent to the node among the plurality of nodes. 2-hop adjacent node address acquisition means for acquiring an address of one adjacent node and an address of a 2-hop adjacent node adjacent to the one adjacent node, and an address of the 2-hop adjacent node for each 2-hop adjacent node 2 hop adjacent node address storage means for storing the address of the adjacent node in association with the 2 hop adjacent address together with the address of the adjacent node when it is determined that the data packet could not be relayed to one of the adjacent nodes One of the addresses of adjacent nodes corresponding to one of the node addresses is the destination. Characterized in that it comprises a packet relay means for relaying packet.

  The wireless node according to the present invention is a wireless node that relays a data packet through a wireless communication path, and is adjacent to the address of the adjacent node and the adjacent node from one of the adjacent nodes adjacent to the wireless node. 2 hop adjacent node address acquisition means for acquiring the address of the 2 hop adjacent node, and 2 hop adjacent for storing the address of the 2 hop adjacent node and the address of the adjacent node in association with each 2 hop adjacent node When it is determined that the data packet could not be relayed to one of the adjacent nodes, the node address storage means and the address of the adjacent node corresponding to one of the addresses of the two-hop adjacent node together with the address of the one adjacent node And packet relay means for relaying the data packet with the destination as the destination.

  According to the packet relay system of the present invention, an appropriate link cost can be calculated even when the link status fluctuates, and an uplink path to the sink node can be established without increasing the amount of communication information in the network. Further, according to the packet relay system of the present invention, it is possible to establish a downstream path from the sink node without causing congestion in the traffic. Also, according to the packet relay system of the present invention, packets can be quickly transmitted to the destination node even when a link break occurs in the downstream path from the sink node.

It is a schematic diagram showing the packet relay system by a 1st Example. It is a block diagram showing the structure of the node by a 1st Example. It is a table | surface which shows an example of a routing table. (A) is a figure showing an example of a format of a sensor data packet. (B) is a figure showing an example of the format of a hello packet. It is a flowchart showing a hello packet transmission process routine. It is a flowchart showing a packet relay process routine. It is a flowchart showing a routing table update process routine. It is a figure showing the link cost between each adjacent node. It is a table | surface which shows an example of the routing table in node b, c, f, e, and h. It is a table | surface which shows an example of a routing table when link cost fluctuates. It is a block diagram showing the structure of the node by a 2nd Example. It is a figure which shows the packet relay system formed in tree shape with the relay direction of a RREC packet. It is a flowchart showing a RREC packet relay processing routine. It is a figure showing the example of aggregation of the RREC packet in each relay step. It is a block diagram showing the structure of the node by a 3rd Example. (A) is a table | surface which shows an example of the routing table of the node c. (B) is a table | surface which shows an example of the 2-hop adjacent node table of the node c. It is a flowchart showing a downlink data packet relay processing routine. It is a figure showing a packet relay system with the address list in each relay step and the relay direction of a downstream data packet. It is a figure showing an example of a format of a downlink data packet. It is a figure showing a packet relay system with the address list in each relay step and the relay direction of a downstream data packet. (A) is a table | surface which shows an example of the routing table of the node c. (B) is a table | surface which shows an example of the 2-hop adjacent node table of the node c. It is a schematic diagram showing the aspect of the packet relay between the nodes at the time of the routing in the conventional many to one system.

  Hereinafter, embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

<First embodiment>
FIG. 1 is a schematic diagram showing a packet relay system 1 according to this embodiment. The packet relay system 1 is a wireless sensor network, for example, and includes a node a that is a sink node and nodes b to j that are sensor nodes. Each of the nodes a to j can exchange a packet by wireless communication with a node adjacent to the node a to j. In the figure, nodes connected by dotted lines can exchange packets with each other. For example, node a can exchange packets with nodes b and f, respectively. In the packet relay system 1, for example, sensor data generated in each of the nodes b to j is aggregated in the node a.

  FIG. 2 is a block diagram showing the configuration of the node b according to this embodiment. Each of the nodes a and c to j has the same configuration as the node b.

  The node b includes an antenna 101, an amplifier 102, a radio signal transmission / reception unit 103, a packet reception unit 104, a relay processing unit 105, a routing table 106, a data packet generation unit 107, a control packet generation unit 108, A packet transmission unit 109.

  The antenna 101 transmits and receives radio signals to and from a node adjacent to the node b.

  The amplifier 102 is an amplifier that amplifies the signal strength of a radio signal transmitted / received by the antenna 101 or an electric signal from the radio signal transmission / reception unit 103.

  The radio signal transmitting / receiving unit 103 performs mutual conversion between the radio signal from the amplifier 102 and the electrical signal from the packet receiving unit 104 or the packet transmitting unit 109, and performs modulation, demodulation, and frequency conversion processing on these signals. The wireless signal is transmitted / received via the amplifier 102 and the antenna 101.

  The packet receiving unit 104 receives a packet included in the wireless signal from the wireless signal transmitting / receiving unit 103.

  When the relay processing unit 105 determines that the packet type is a control packet based on the information included in the packet received by the packet receiving unit 104, the relay processing unit 105 determines that the packet is a control packet. The routing table 106 is updated based on the information included in. If the relay processing unit 105 determines that the packet is a data packet, the relay processing unit 105 determines a relay destination node based on information in the routing table. The relay processing unit 105 is an arithmetic circuit such as a microprocessor, for example.

  The routing table 106 stores a link cost, a path cost, and a packet arrival rate for a node adjacent to the node b. The routing table 106 is a memory such as a RAM.

  FIG. 3 is a table showing an example of the routing table 106 of the node b. The link cost, path cost, and packet arrival rate for each of the nodes a, f, c, and e adjacent to the node b are stored. For example, the link cost for the adjacent node a is 2, the path cost is 2, and the packet arrival rate is 0.8. These pieces of information are updated by the relay processing unit 105.

  The data packet generator 107 generates sensor data corresponding to a sensor signal from a sensor (not shown), and generates a sensor data packet including the sensor data.

  The control packet generation unit 108 sets a packet arrival rate, a link cost, and a path cost from a node adjacent to the node b (hereinafter referred to as a hello packet), and a downlink path setting from the node a. A control packet such as an RREC packet is generated. The control packet generator 108 and the data packet generator 107 are arithmetic circuits such as a microprocessor, for example.

  FIG. 4A is a diagram illustrating an example of a format of the sensor data packet. FIG. 4B is a diagram illustrating an example of a format of a hello packet.

  The header part of the sensor data packet and the hello packet includes a packet type that is an identifier indicating the type of the packet, a source address that is the address of the packet transmission source node, a relayable number of times that indicates a limit on the number of packet relays, and each packet A sequence number which is a serial number attached to is included.

  The payload portion of the sensor data packet includes packet length, number of retransmissions, for example, data such as sensor data. The payload portion of the hello packet includes a packet length and a path cost, and may further include a link cost for each adjacent address. Here, the path cost is an accumulated value of link costs (link cost accumulated value) calculated by each node from the node a which is a sink node to its own node.

  The packet transmission unit 109 causes the wireless signal transmission / reception unit 103 to transmit the packet generated by the data packet generation unit 107 or the control packet generation unit 108 to a node adjacent to the node b.

  Each of the nodes a to j roughly executes two processes of a hello packet transmission process and a packet relay process in parallel.

  FIG. 5 is a flowchart showing a hello packet transmission processing routine. The control packet generator 108 determines whether or not a predetermined periodical hello packet transmission timing such as an interval of several seconds to several tens of seconds has arrived (step S101), and when the timing arrives, the hello packet Is generated (step S102). At this time, the control packet generation unit 108 generates a hello packet in which the minimum path cost among path costs for each adjacent node stored in the routing table 106 is included in the payload part. For example, when the routing table 106 at the time of generating the hello packet in the node b is as shown in FIG. 3, the minimum path cost “2” is included in the payload portion. The packet transmission unit 109 causes the wireless signal transmission / reception unit 103 to transmit the hello packet generated by the control packet generation unit 108 to a node adjacent to the node b (step S103).

  FIG. 6 is a flowchart showing a packet relay processing routine in each of the nodes a to j. FIG. 7 is a flowchart showing a routing table update processing routine included in the packet relay processing routine. FIG. 8 is a diagram showing the link cost between adjacent nodes in the nodes a to j by numerical values between the adjacent nodes. FIG. 9 is a table showing an example of the routing table 106 in the nodes b, f, c, e, and h. Hereinafter, the packet relay process will be described with reference to FIGS.

  First, the packet relay process in the node b adjacent to the node a which is the sink node will be described in order.

  The relay processing unit 105 of the node b determines whether or not the packet is a hello packet from the packet type included in the header part of the packet received by the packet receiving unit 104 (step S201).

  When the relay processing unit 105 determines that the packet is a hello packet, the relay processing unit 105 updates the routing table 106 (step S202). The format of the hello packet is as shown in FIG. 4B as described above.

  First, the relay processing unit 105 determines the node that transmitted the packet from the source address included in the header of the packet, and calculates the packet arrival rate for the node (step S301). When the relay processing unit 105 determines that the packet is from, for example, the node a, the relay processing unit 105 calculates the packet arrival rate from the reception history of the hello packet from the node a and the sequence number included in the header part of the packet. For example, when the number of hello packets from the node a received up to the time of calculating the packet arrival rate is 80 and the sequence number of the packet is 100, the packet arrival rate is 0.8 (= 80/100). . The relay processing unit 105 updates the packet arrival rate for the node a obtained by the calculation (FIG. 9B). If the source address of the packet is not recorded in the routing table 106, the source address is newly added.

  Next, the relay processing unit 105 calculates a link cost for each adjacent node (step S302). The relay processing unit 105 calculates, for example, link cost = 10−packet arrival rate × 10. For example, when the packet arrival rate for node a is 0.8, link cost = 10−0.8 × 10 = 2. The relay processing unit 105 updates the link cost for the node a obtained by the calculation (FIG. 9B).

  Next, the relay processing unit 105 acquires the path cost (link cost accumulated value) included in the packet (step S303). For example, if the packet is from node a, the path cost is 0 because node a is a sink node.

  Next, the relay processing unit 105 calculates a path cost from the node b to the node a when the packet is relayed to the node a (step S304). Specifically, the relay processing unit 105 adds the link cost “2” for the node a stored in the routing table 106 and the path cost “0” acquired from the packet from the node a to add the path cost. “2” = (2 + 0) is calculated. The relay processing unit 105 updates the path cost for the node a obtained by the calculation (FIG. 9B).

  Similarly to the node b, in the case of the node f adjacent to the node a, the packet arrival rate, the link cost and the path cost for the node a are calculated in the same manner and the routing table 106 is updated (FIG. 9 (f)).

  Since node b and node f are also adjacent to each other, hello packets are exchanged with each other.

  First, the node b calculates the packet arrival rate “0.5” of the hello packet from the node f (step S301), and calculates the link cost “5” for the node f (step S302). Since the node f transmits the hello packet by including “3”, which is the minimum path cost in its own routing table 106, in the payload portion, the node b sets the path cost “3” of the hello packet from the node f. Obtaining (step S303), the path cost from node b to node a when the packet is relayed to node f is calculated (step S304). Specifically, the node b adds the link cost “5” for the node f stored in the routing table 106 and the path cost “3” acquired from the packet from the node f to add the path cost “8”. ] = (5 + 3) is calculated. The relay processing unit 105 of the node b updates the packet arrival rate, link cost, and path cost for the node f obtained by the calculation (FIG. 9B).

  Similarly, in the case of the node f, the packet arrival rate, link cost and path cost for the node b are calculated and the routing table 106 is updated (FIG. 9 (f)).

  Next, the case of the node c adjacent to the node b will be described. First, the node c calculates the packet arrival rate “0.9” of the hello packet from the node b (step S301), and calculates the link cost “1” for the node b (step S302). Since the node b transmits the hello packet including “2”, which is the minimum path cost in its own routing table 106, in the payload portion, the node c sets the path cost “2” of the hello packet from the node b. Obtaining (step S303), the path cost from node c to node a when the packet is relayed to node b is calculated (step S304). Specifically, the node c adds the link cost “1” for the node b stored in the routing table 106 and the path cost “2” acquired from the packet from the node b to add the path cost “3”. "= (1 + 2) is calculated. The relay processing unit 105 of the node c updates the packet arrival rate, link cost, and path cost for the node b obtained by the calculation (FIG. 9C).

  Similarly, in the case of the node e adjacent to the nodes b and f, the packet arrival rate, the link cost, and the path cost for the nodes b and f are calculated in the same manner to update the routing table 106 (FIG. 9E).

  Since node c and node e are also adjacent to each other, they exchange hello packets with each other.

  First, the node c calculates the packet arrival rate “0.6” of the hello packet from the node e (step S301), and calculates the link cost “4” for the node e (step S302). Since the node e transmits the hello packet by including “4”, which is the minimum path cost in its own routing table 106, in the payload portion, the node c adds the path cost “4” of the hello packet from the node e. Obtaining (step S303), the path cost from node c to node a when the packet is relayed to node e is calculated (step S304). Specifically, the node c adds the link cost “4” for the node e stored in the routing table 106 and the path cost “4” acquired from the packet from the node f to add the path cost “8”. ] = (4 + 4) is calculated. The relay processing unit 105 of the node c updates the packet arrival rate, link cost, and path cost for the node e obtained by the calculation (FIG. 9C).

  Similarly, in the case of the node e, the packet arrival rate, link cost, and path cost for the node c are calculated and the routing table 106 is updated (FIG. 9 (e)).

  Next, the case of the node h located farthest from the sink node a will be described. The node h calculates the packet arrival rate and the link cost for each of the nodes d, g, and j by the same processing as described above according to the hello packet from each of the adjacent nodes d, g, and j ( Steps S301 and S302).

  Each of the nodes d, g, and j also calculates the link cost and path cost for each adjacent node and stores them in the routing table 106 in the same manner as described above. Node d is the minimum path cost from node d to node a [6] (path via nodes a, b and c), and node g is the minimum path cost from node g to node a [5] (node node j), and node j transmits the minimum path cost [6] from node j to node a (path via nodes a, f and i) in each hello packet. . The node h adds the path cost “6” acquired from the hello packet from the node d and the link cost “2” for the node d stored in the routing table 106 to obtain the path cost “8” = (6 + 2 ) Is calculated. Similarly, the node h calculates path costs “6” and “7” for each of the nodes g and j.

  In this manner, each of the nodes b to j updates the routing table 106 by exchanging hello packets between adjacent nodes.

  With reference to FIG. 6 again, the packet relay processing in the node h will be described.

  When the relay processing unit 105 of the node h determines that the packet is an uplink data packet from the packet type included in the header portion of the packet received by the packet receiving unit 104 (step S203), the relay destination node of the packet Is selected (step S204). As described above, the format of the data packet is as shown in FIG.

  At this time, the relay processing unit 105 selects the node g corresponding to the minimum path cost “6” in the routing table 106 (FIG. 9H) as the relay destination node, and performs the radio signal transmitting / receiving unit 103 to g to transmit the data packet (step S205). In this case, the data packet is transferred in the order of nodes h, g, c, b, and a. At this time, the data packet can be relayed at the minimum path cost “6” from the node h to the node a.

  When the packet reception unit 104 receives the reception confirmation packet from the node g, the relay processing unit 105 determines that the packet relay has been successful (step S206), and ends the packet relay processing.

  On the other hand, if the reception confirmation packet from the node g is not received by the packet receiving unit 104, the relay processing unit 105 determines that the packet relay has failed (step S206), and determines the relay destination node again. Select (step S204). In this case, the relay processing unit 105 selects the node j corresponding to the path cost “7” next to the path cost “6” corresponding to the node g as the relay destination node. The relay processing unit 105 causes the wireless signal transmission / reception unit 103 to transmit the data packet to the node j (step S205). In this case, the data packet is transferred in the order of the nodes h, j, i, f, and a.

  If the relay processing unit 105 determines that the relay of the packet to the node j has also failed, the relay processing unit 105 sets the node d corresponding to the path cost “8” next to the node j corresponding to the node j to the relay destination node. The wireless signal transmitting / receiving unit 103 transmits the data packet to the node d (step S205). If it is determined that packet relay has failed for any of the nodes d, g, and j, the node g corresponding to the minimum path cost in the routing table 106 is selected again as a relay destination node. The packet relay process is continued, and the process ends when it is determined that the packet relay is successful.

  Further, for example, when a link situation between these two nodes deteriorates due to passage of a shield such as a person or a car between the node h and the node g, the arrival rate of the hello packet from the node g to the node h is increased. Getting worse. For example, when the arrival rate deteriorates to 0.7, the link cost for the node g calculated by the node h is 3, and the path cost is 8. In this case, the routing table 106 of the node h is updated as shown in FIG. That is, since the minimum path cost in the routing table 106 is 7, the relay processing unit 105 selects the node j corresponding to the minimum path cost “7” in the routing table 106 (FIG. 10) as the relay destination node. Then, the wireless signal transmission / reception unit 103 is made to transmit the data packet to the node j. In this case, the data packet is transferred in the order of the nodes h, j, i, f, and a.

  When relaying the data packet, the relay processing unit 105 subtracts the value of the relayable number included in the header part. If the relay processing unit 105 determines that the value of the number of possible relays is 0, the relay processing unit 105 ends the packet relay process without relaying the data packet. Each time the relay processing unit 105 determines that the data packet relay has failed, the relay processing unit 105 subtracts the value of the number of retransmissions included in the payload portion. When determining that the value of the number of retransmissions is 0, the relay processing unit 105 ends the packet relay process without relaying the data packet.

  When the relay processing unit 105 determines that the packet is not an uplink data packet from the packet type included in the header portion of the packet received by the packet receiving unit 104 (step S203), the relay processing unit 105 proceeds to the RREC packet relay processing routine. (Step S207). The RREC packet relay process will be described in the second embodiment. Each of the nodes b to g, i, and j also performs data packet relay processing in the same manner as described above.

  As described above, according to the packet relay system according to the present embodiment, each of the nodes b to j can select an optimum route from the node b to the node a. That is, an adjacent node to be a data packet relay destination node can be selected so that the path cost from itself to the node a is minimized.

  Each node exchanges a packet with a node adjacent to itself, and calculates a link cost for the corresponding adjacent node based on the arrival rate. Therefore, an appropriate link cost can be calculated even when the link status fluctuates between adjacent nodes. In addition, since the path cost to the sink node is calculated by adding each link cost from itself to the sink node, it is possible to calculate an appropriate value according to the change in the link status for the path cost.

  In addition, each node obtains the accumulated value of the link cost calculated by each node from the node not adjacent to itself, that is, from the sink node to the adjacent node, from the adjacent node, and the accumulated value and the adjacent value calculated by itself The path cost to the sink node is calculated from the link cost for the node. Since the path cost is calculated by such processing, the upstream path to the sink node can be established without increasing the amount of communication information in the network.

  In addition, since each node calculates and stores the path cost for each adjacent node, even when packet relay to the optimal adjacent node fails, the adjacent node to be the next packet relay target is appropriately selected. be able to.

<Second embodiment>
FIG. 11 is a block diagram illustrating the configuration of the node b according to the present embodiment. Each of the nodes a and c to j has the same configuration. The packet aggregation unit 110 is further included in the configuration of the node b in the first embodiment.

  The packet aggregating unit 110 holds the packet until a predetermined transmission timing arrives when the relay processing unit 105 determines that the RREC packet from the adjacent node 1 has been received, and performs relay processing until the transmission timing arrives. When the unit 105 determines that an RREC packet from an adjacent node other than the one adjacent node has been received, a new RREC packet (hereinafter also referred to as an aggregate packet) in which the addresses included in both packets are aggregated The packet is transmitted to the adjacent node of the relay destination by the packet transmission unit 109 according to the arrival of the transmission timing.

  FIG. 12 is a diagram illustrating the packet relay system 2 formed in a tree shape together with the relay direction of the RREC packet. The packet relay system 2 includes a node a which is the highest node and nodes b to h which are nodes below the node a and are arranged in a tree shape while forming a hierarchical structure. Each of the nodes b to h sequentially relays the RREC packet from the node located on the lower side to the node located on the upper side as indicated by an arrow in FIG. Each of the nodes b to h relays the RREC packet by including its own address in the RREC packet. The packet aggregating unit 110 generates a new RREC packet in which these addresses included in the RREC packet are aggregated.

  FIG. 13 is a flowchart showing the RREC packet relay processing routine. This RREC packet relay process routine is a process corresponding to the RREC packet relay process in step S207 in the packet relay process routine shown in FIG. FIG. 14 is a diagram illustrating an example of aggregation of RREC packets at each relay stage, corresponding to the numbers shown with arrows in FIG. Hereinafter, the RREC packet relay process will be described with reference to FIGS.

  First, the RREC packet relay processing in the lowest node, node h, will be described in order.

  The control packet generation unit 108 of the node h generates an RREC packet including the address of the node h and transmits it to the node d adjacent to the node h ((1) in FIG. 12). At this time, as shown in FIG. 14A, the control packet generation unit 108, for example, the header HD, the packet size LN of the payload portion, the number of addresses NM “1” included in the packet, An RREC packet including an address AD “h” such as an IP address is generated. Hereinafter, the address of the node h is simply referred to as an address h. Similarly, addresses of other nodes are referred to as address a, address b,..., Address g.

  The relay processing unit 105 of the node d determines whether or not the packet is an RREC packet from the packet type included in the header part of the packet received by the packet receiving unit 104 (step S401). When the relay processing unit 105 determines that the packet is an RREC packet, the packet aggregation unit 110 holds the packet information included in the packet until a predetermined transmission timing arrives (steps S402 and S403). For example, when the packet is an RREC packet from the node h, the packet aggregation unit 110 holds information such as an address h included in the RREC packet. The predetermined transmission timing is a periodic timing such as several seconds to several tens of seconds, for example, and is a preset timing.

  When the predetermined transmission timing has arrived, the packet aggregating unit 110 of the node d transmits the RREC packet including the addresses d and h to the higher-level node node b (step S405). At this time, the packet aggregating unit 110 generates an RREC packet including the header HD and the packet size LN, the number of addresses NM “2”, and the addresses d and h, as shown in FIG. Since the packet receiving unit 104 of the node d does not receive RREC packets from nodes other than the node h, the packet aggregating unit 110 omits the RREC packet aggregation processing in step S404.

  The lower node e of the node b transmits the RREC packet as shown in FIG. 14 (3) to the upper node node b by the same processing as the node h.

  When the relay processing unit 105 determines that the packet received by the packet reception unit 104 is an RREC packet (step S401), the packet aggregation unit 110 of the node b converts the packet information included in the RREC packet to a predetermined value. This is held until the transmission timing arrives (steps S402 and S403). For example, when the packet is an RREC packet from the node d, the packet aggregation unit 110 holds the addresses d and h included in the RREC packet.

  If the packet aggregating unit 110 of the node b determines that the RREC packet has been further received by the packet receiving unit 104 before the relay processing unit 105 arrives at a predetermined transmission timing (step S401), the packet aggregating unit 110 in the same way The included packet information is held until a predetermined transmission timing arrives (steps S402 and S403). For example, when the packet is an RREC packet from the node e, the packet aggregation unit 110 holds the address e included in the RREC packet.

  When the predetermined transmission timing has arrived (step S403), the packet aggregation unit 110 of the node b aggregates the stored address information and generates a new RREC packet (step S404). As shown in FIG. 14 (4), the packet aggregation unit 110 includes a header HD and a packet size LN, an address group composed of addresses of a plurality of nodes having a parent-child relationship, and the number of addresses for each address group. Generate a RREC packet.

  At this time, the packet aggregating unit 110 of the node b uses the address b of the node b that is its own node as the head address, the address d of the node d that is a child node of the node b as the second address, and also a child of the node b. An address group (addresses b, d, e) having the address e of the node e as a third address is generated, and the number NM is set to “3”. Further, the packet aggregating unit 110 generates an address group (address d, h) in which the address d of the node d is the first address and the address h of the node h that is a child node of the node d is the second address. The number NM is set to “2”. In this way, the packet aggregating unit 110 generates a new RREC packet (aggregated packet) as shown in FIG. 14 (4).

  The packet aggregating unit 110 transmits a new RREC packet (FIG. 14 (4)) to the node a that is the upper node (step S405).

  The node c performs the same RREC packet relay processing as the node b. The packet aggregating unit 110 of the node c obtains the address f included in the RREC packet (FIG. 14 (5)) from the node f and the address g included in the RREC packet (FIG. 14 (6)) from the node g. This is held until a predetermined transmission timing comes (step S402), and when the transmission timing comes (step S403), a new RREC packet is generated.

  At this time, the packet aggregating unit 110 of the node c uses the address c of the node c that is its own node as the head address, the address f of the node f that is a child node of the node c as the second address, and also a child of the node c. An address group (addresses c, f, g) having the address g of the node g as a third address is generated, the number NM is set to “3”, and a new RREC as shown in FIG. Generate packets (aggregated packets). The packet aggregating unit 110 transmits a new RREC packet ((7) in FIG. 14) to the node a that is the upper node (step S405).

  The node a receives the RREC packet (FIG. 14 (4) and (7)) from each of the nodes b and c which are lower nodes. When the node a is a sink node, the RREC packet is not relayed. However, when the node a relays the RREC packet to a higher-order node, the packet aggregating unit 110 receives the RREC from each of the nodes b and c. The addresses included in the packets are aggregated to generate an RREC packet as shown in FIG.

  At this time, the packet aggregating unit 110 of the node “a” uses the address “a” of the node “a” as its own node as the head address, the address “b” of the node “b” as the child node of the node “a” as the second address, and An address group (addresses a, b, c) having the address c of the node c as a third address is generated, and the number NM is set to “3”. Further, the packet aggregating unit 110 uses the address b of the node b as the head address, the address d of the node d as a child node of the node b as the second address, and the address e of the node e as a child node of the node b. An address group (addresses b, d, e) as a third address is generated, and the number NM is set to “3”.

  Similarly, the packet aggregating unit 110 generates an address group (address d, h) having an address number NM “2” and an address group (address c, f, g) having an address number NM “3”. A new RREC packet (FIG. 14 (8)) including all address groups is generated. The packet aggregating unit 110 transmits an RREC packet as shown in FIG. 14 (8) to a higher-order node (not shown).

  As described above, according to the packet relay system according to the present embodiment, when the RREC packet is sequentially relayed from the lower node to the upper node among the nodes formed in the tree shape, the RREC packet from each lower node is used. A new RREC packet (aggregated packet) generated by aggregating the addresses included in the address group collected for each parent-child relationship is transmitted to the upper node.

  In the conventional packet relay system, the relay node relays the RREC packets transmitted from each lower node by unicast transmission as they are without being aggregated. Therefore, when the transmission timing in each node overlaps, However, according to the packet relay system of this embodiment, the addresses included in the RREC packets from the lower nodes are aggregated and one RREC packet is sent to the upper node. Since it is relayed, there is no congestion in the traffic. In addition, since each address group included in the RREC packet is grouped for each parent-child relationship, the highest-level sink node can grasp the connection relationship of each lower-level node, and the downlink from the sink node A route can be established.

<Third embodiment>
FIG. 15 is a block diagram illustrating the configuration of the node c according to this embodiment. Each of the nodes a, b, and d to j has the same configuration. The configuration of the node b in the first embodiment further includes a 2-hop adjacency table 111.

  FIG. 16A is a diagram illustrating an example of the routing table 106 of the node c. FIG. 16B is a diagram illustrating an example of the 2-hop adjacent node table 111 of the node c. The node c has a 2-hop adjacency table 111 in addition to the routing table 106 similar to that in the first embodiment. The 2-hop adjacency table 111 stores a 1-hop node address and a 2-hop node address. Here, the 1-hop node refers to a node adjacent to the node c, and the 2-hop node refers to a node adjacent to the 1-hop node and not adjacent to the node c. . For example, the 1-hop nodes for node c are nodes b, e, d, and g, and the 2-hop nodes are nodes a, f, e, j, and h.

  The relay processing unit 105 of the node c receives the adjacent address included in the payload portion of the hello packet as shown in FIG. Based on this, the addresses of the 1-hop and 2-hop nodes are stored in the 2-hop adjacency table 111. For example, since the node b is adjacent to the nodes a, f, and e in addition to the node c, the hello packet having addresses a, f, and e as the adjacent addresses of the payload portion is transmitted to the node c.

  The relay processing unit 105 of the node c stores the addresses that are not adjacent to the node c among the adjacent addresses, that is, the addresses a and f in the 2-hop adjacency table 111 as the addresses of the 2-hop nodes, and the addresses a and f Address b is stored in association with each. When the relay processing unit 105 receives a hello packet from another adjacent node, the relay processing unit 105 performs the same processing to generate a 2-hop adjacent table 111 as shown in FIG.

  FIG. 17 is a flowchart showing a downstream data packet relay processing routine. This downlink data packet relay process routine corresponds to the downlink data packet relay process in step S406 in the RREC packet relay process routine shown in FIG. FIG. 18 is a diagram illustrating the packet relay system 1 together with the relay direction of the downlink data packet and the address list at each relay stage. Hereinafter, the downlink data packet relay processing from the node a serving as the sink node to the node h will be described with reference to FIGS. 17 and 18.

  First, the processing in the node a which is a sink node will be described. FIG. 19 is a diagram illustrating an example of a format of a downlink data packet. In the header part of the data packet, a packet type that is an identifier indicating the type of the packet, a source address that is the address of the packet transmission source node, a relayable number that indicates a limit on the number of packet relays, and a packet-specific header Contains a sequence number that is a sequential number.

  The payload portion of the downlink data packet includes a packet length, the number of addresses, addresses corresponding to the number of addresses, and data to be transmitted to the destination. Hereinafter, a group of addresses included in the payload portion is referred to as an address list. In the address list of the downlink data packet transmitted by the packet transmission unit 109 of the node a, addresses are described in the order of addresses b, c, g, and h. This order is set by the relay processing unit 105 based on the node connection status obtained from the address group included in the RREC packet from the lower node received by the node a as described in the second embodiment. In order.

  The packet transmission unit 109 of the node a transmits the downlink data packet to the node b, that is, the address b described at the top of the address list.

  The relay processing unit 105 of the node b changes the address list by deleting the address b (that is, the address of its own node) described at the head of the address list included in the downlink data packet from the node a (step S501). ). The relay processing unit 105 transmits the downlink data packet including the changed address list to the node c, that is, the address c described at the top of the address list (step S502).

  When the relay processing unit 105 of the node b determines that the packet receiving unit 104 has received the packet reception notification from the node c, the relay processing unit 105 determines that the packet relay is successful (step S503), and ends the relay processing. When it is determined that the packet reception unit 104 has not received the packet reception notification from the node c, the relay processing unit 105 of the node b determines that the packet relay has failed (step S503) and stores the 2-hop adjacency table 111. Reference is made (step S504). Here, it is assumed that the packet relay is successful, and the relay processing unit 105 ends the packet relay process.

  The relay processing unit 105 of the node c changes the address list by deleting the address c (that is, the address of its own node) described at the top of the address list included in the downlink data packet from the node b (step S501). ). The relay processing unit 105 relays the downlink data packet including the changed address list to the node g, that is, the address g described at the top of the address list (step S502). Here, it is assumed that a downlink data packet cannot reach node g due to an influence of an instantaneous link break caused by, for example, a shield such as a person or a car passing between node c and node g.

  When the relay processing unit 105 of the node c determines that the packet receiving unit 104 has not received the packet reception notification from the node g, that is, the packet relay has failed (step S503), the relay processing unit 105 of FIG. The hop adjacency table 111 is referred to (step S504). First, the relay processing unit 105 searches whether the address h described next to the address g in the address list is stored as the address of the 2-hop node in the 2-hop adjacency table 111 (step S505).

  When the address h is stored as the address of the 2-hop node, the relay processing unit 105 downloads the address of the unrelayed node, that is, the address d, among the addresses d and g of the 1-hop node corresponding to the address h. Determine the destination of the data packet. The relay processing unit 105 changes the address in the address list to addresses d and h (step S501), and relays to the node d (step S502).

  The relay processing unit 105 of the node c determines that the packet reception unit 104 has received a packet reception notification from the node d (step S503), and ends the packet relay processing. The node d also relays the downlink data packet whose address list is changed to the address h to the node h by the same process as described above.

  As described above, according to the packet relay system according to the present embodiment, each node stores the address of the node that is two hops ahead included in the packet exchanged with the adjacent node in association with the address of the adjacent node. When the node determines that packet relay to one of the adjacent nodes has failed, it selects 1 of the address associated with the address of the relay destination node that is two hops ahead, and the adjacent node of the one address Relay the packet to

  In the conventional packet relay system, when the packet is not correctly transferred to the destination node, the RREC command is transmitted from the destination node to the sink node in order to reestablish the downlink route from the sink node. There is a problem in that downlink communication cannot be performed until the downlink route is re-established. On the other hand, according to the relay system according to the present embodiment, the packet relay destination node is changed and the packet is relayed. Therefore, even when the packet relay fails on the downstream path from the sink node, the packet is sent to the destination node. Can be relayed quickly.

  The above example is an example in the case where the node c is the only alternative node candidate when the node c fails in packet relay. However, when there are a plurality of alternative node candidates, the following processing is performed.

  FIG. 20 is a diagram illustrating the packet relay system 1 together with the relay direction of the downlink data packet and the address list at each relay stage. A node k is further added to the packet relay system 1 in the above example. Node k is a node adjacent to nodes c and h, as shown in FIG. FIG. 21A is a diagram illustrating an example of the routing table of the node c in this case. An adjacent address, link cost, path cost, and arrival rate for node k are added. FIG. 21B is a diagram illustrating an example of the 2-hop adjacent node table of the node c in this case. Addresses d, g, and k are associated with the address h of the 2-hop node h.

  When the relay processing unit 105 of the node c determines that the downlink data packet relay to the node g has failed (step S503), the relay processing unit 105 refers to the 2-hop adjacency table 111 as illustrated in FIG. 21B (step S504). ).

  The relay processing unit 105 searches whether the address h described next to the address g in the address list is stored as the address of the 2-hop node in the 2-hop adjacency table 111 (step S505).

  When the address h is stored as the address of a 2-hop node, the relay processing unit 105 transmits the address of an unrelayed node among the addresses d and g of the 1-hop node corresponding to the address h, that is, the address d or k. Is determined as the destination of the downstream data packet.

  At this time, the relay processing unit 105 refers to the routing table as shown in FIG. 21A and compares the link cost “3” corresponding to the address d with the link cost “1” corresponding to the address k. The address k having the smaller link cost is determined as the relay destination of the downlink data packet. The relay processing unit 105 changes the address in the address list to the addresses k and h (step S501), and relays to the node k (step S502). By such processing, a more appropriate relay destination node can be selected.

  In addition, the above example is an example in the case where the reception confirmation response from the relay destination node is not obtained even once, or another relay destination node is selected immediately, but the reception confirmation response is not obtained. However, packet transmission may be performed a plurality of times to the same relay destination node, and another relay destination node may be selected when it is determined that all of the plurality of transmissions have failed.

1 packet relay system 101 antenna 102 amplifier 103 wireless signal transmission / reception unit 104 packet reception unit 105 relay processing unit 106 routing table 107 data packet generation unit 108 control packet generation unit 109 packet transmission unit 110 packet aggregation unit 111 2-hop adjacency table

Claims (4)

A packet relay system including a plurality of nodes that relay data packets to each other,
Each of the nodes is
2-hop adjacent node address acquisition means for acquiring the address of the one adjacent node and the address of the 2-hop adjacent node adjacent to the one adjacent node from one of the adjacent nodes adjacent to the node among the plurality of nodes; ,
2-hop adjacent node address storage means for storing the address of the 2-hop adjacent node and the address of the adjacent node in association with each other for each 2-hop adjacent node;
When it is determined that the data packet could not be relayed to one of the adjacent nodes, the data packet is addressed to one of the addresses of the adjacent node corresponding to one of the addresses of the two-hop adjacent node together with the address of the one adjacent node A packet relay system for relaying the packet. The node is
Arrival rate calculation means for calculating the arrival rate of packets from each of the adjacent nodes for each of the adjacent nodes;
Link cost calculating means for calculating a link cost for each of the adjacent nodes based on the reach rate;
When the packet relay means determines that the data packet could not be relayed to one of the adjacent nodes, the address of the adjacent node corresponding to one of the addresses of the two-hop adjacent node together with the address of the one adjacent node The packet relay system according to claim 1, wherein the data packet is relayed with an address corresponding to an adjacent node having the smallest link cost as a destination. A wireless node that relays data packets via a wireless communication path,
2-hop adjacent node address acquisition means for acquiring an address of the one adjacent node and an address of a 2-hop adjacent node adjacent to the one adjacent node from one of adjacent nodes adjacent to the node;
2-hop adjacent node address storage means for storing the address of the 2-hop adjacent node and the address of the adjacent node in association with each other for each 2-hop adjacent node;
When it is determined that the data packet could not be relayed to one of the adjacent nodes, the data packet is addressed to one of the addresses of the adjacent node corresponding to one of the addresses of the two-hop adjacent node together with the address of the one adjacent node And a packet relay means for relaying the wireless node. Arrival rate calculation means for calculating the arrival rate of packets from each of the adjacent nodes for each of the adjacent nodes;
Link cost calculating means for calculating a link cost for each of the adjacent nodes based on the reach rate;
When the packet relay means determines that the data packet could not be relayed to one of the adjacent nodes, the address of the adjacent node corresponding to one of the addresses of the two-hop adjacent node together with the address of the one adjacent node The radio node according to claim 3, wherein the data packet is relayed with an address corresponding to an adjacent node having the smallest link cost as a destination.

Images