JP2016201654A - Communication device and relaying method of packet - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve the throughput of a communication network, by reducing the latency of packet transfer.SOLUTION: A communication device 10 is used in a communication node N performing the routing control in a communication network NW autonomously. The communication device 10 includes a storage unit 12 for storing the path entry of each destination node where the effective period is set, and a control unit 11 for relaying a packet, addressed to other station, to an adjacent node. The relay processing performed by the control unit 11 includes transfer processing of a packet carried out first, and update processing for updating the effective period of a path entry corresponding to the packet to a set value.SELECTED DRAWING: Figure 3

Description

The present invention relates to a communication device and a packet relay method.
Specifically, the present invention relates to a technique for reducing packet transfer latency in a communication device used for a communication node that autonomously performs path control in a communication network.

As a communication network in which each communication node autonomously searches for a route, there is a communication network of a distance vector type routing protocol. The distance vector type routing protocol is an algorithm for determining a suitable route based on distance (distance) and vector (direction).
For example, AODV (Adhoc ON-Demand Distance Vector) is a typical example of a distance vector type routing protocol.

In AODV, each communication node holds a route entry for each destination node, and each communication node determines information to be written in its own route entry when receiving a route request (RREQ) and a route response (RREP). As a result, a path between predetermined communication nodes is established.
The route entry for each destination node held by each communication node includes a lifetime (Lifetime) that is updated to a predetermined setting value (recommended is 3 seconds) each time the route entry is used. (See Non-Patent Document 1).

RFC3561 2. Overview

In a conventional communication apparatus that implements AODV, when a packet addressed to another station is received, the valid period of the route entry is updated before the packet is transmitted, and the packet is transmitted after the update is completed. .
The reason is considered that if a packet is not relayed within the effective period of the route entry, the route entry becomes invalid and a new route search is required, so priority is given to securing the effective period of the route entry.

However, in the relay process including the update of the effective period between the reception and transmission of the packet, the packet transfer latency increases by the amount including the update of the effective period.
As a result, the throughput of data communication is lowered, and there is a problem that the communication speed of the entire network is deteriorated. This problem becomes more prominent as the number of communication nodes included in the communication network increases.

  SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems, and an object of the present invention is to provide a communication apparatus and the like that can improve the throughput of a communication network by reducing packet transfer latency.

  (1) A communication apparatus according to an aspect of the present invention is a communication apparatus used for a communication node that autonomously performs path control in a communication network, and stores a path entry for each destination node for which a valid period is set. A storage unit, and a control unit that performs relay processing for relaying a packet addressed to another station corresponding to the route entry during the effective period to an adjacent node, and the relay processing performed by the control unit includes: The packet transfer process performed first and the update process performed after the transfer process for updating the valid period of the route entry corresponding to the packet to a set value are included.

  (4) A relay method according to an aspect of the present invention is a packet relay method performed by a communication device used for a communication node that autonomously performs route control in a communication network, and for each destination node for which a valid period is set And a second step of performing a relay process of relaying a packet addressed to another station corresponding to the route entry during the effective period to an adjacent node. The relay process includes a transfer process of the packet that is performed first and an update process that is performed after the transfer process and updates the valid period of the route entry corresponding to the packet to a set value. .

  According to the present invention, it is possible to improve the throughput of a communication network by reducing packet transfer latency.

It is a figure which shows the connection form of the communication system which concerns on embodiment of this invention. It is a block diagram which shows the internal structure of a communication apparatus. It is a block diagram which shows the internal structure of the control part of a communication apparatus. It is explanatory drawing which shows the conventional packet relay process. It is explanatory drawing which shows the packet relay process of 1st Embodiment. It is explanatory drawing which shows the packet relay process of 2nd Embodiment.

<Outline of Embodiment of the Present Invention>
Hereinafter, an outline of embodiments of the present invention will be listed and described.
(1) A communication apparatus according to the present embodiment is a communication apparatus used for a communication node that autonomously performs path control in a communication network, and stores a path entry for each destination node for which a valid period is set; A relay unit that relays a packet addressed to another station corresponding to the route entry during the valid period to an adjacent node, and the relay unit performed by the controller includes And the update process for updating the valid period of the route entry corresponding to the packet to a set value, which is performed after the transfer process.

According to the communication apparatus of this embodiment, in the relay process of a packet addressed to another station, the control unit first performs the packet transfer process and updates the valid period of the route entry corresponding to the packet to the set value. Therefore, the packet transfer latency can be reduced by the time required for updating the valid period, compared to the conventional relay process.
For this reason, by adopting the above communication device as a communication node, the packet transfer latency of each communication node is reduced and the throughput of the communication network is improved.

(2) In the communication apparatus according to the present embodiment, the update process includes, for example, a process of accumulating address information of a plurality of transferred packets and a plurality of address information of the accumulated packets. And a process of updating the valid period of the route entry corresponding to the packet to a set value.
In this case, there is an advantage that the execution period of the packet transfer process can be shortened and the throughput of the communication network can be further improved as compared with the case where the valid period is updated every time the packet transfer completion is detected.

(3) In the communication apparatus according to the present embodiment, the update process includes a process for detecting completion of transfer of the packet, and the valid period of the route entry corresponding to the packet every time detection of transfer completion of the packet is detected. And a process of updating to a set value.
In this case, there is no need to provide a buffer for storing address information of a plurality of packets, so that there is an advantage that it can be implemented more easily than when the buffer is provided.

  (4) The packet relay method of the present embodiment is a packet relay method performed by a communication device used for a communication node that autonomously performs route control in a communication network, and a valid period is set. A first step of storing a route entry for each destination node, and a second step of performing a relay process for relaying a packet addressed to another station corresponding to the route entry during the validity period to an adjacent node, The relay process of the second step includes the transfer process of the packet that is performed first, and the update process that is performed after the transfer process and updates the valid period of the route entry corresponding to the packet to a set value. , Is included.

According to the relay method of this embodiment, in the relay process of the second step, the packet transfer process is performed first, and the update process for updating the valid period of the route entry corresponding to the packet to the set value is performed later. Compared with the relay processing of the above, the packet transfer latency can be reduced by the time required for updating the valid period.
For this reason, by employing a communication device that performs the above-described relay method as a communication node, the packet transfer latency of each communication node is reduced and the throughput of the communication network is improved.

<Details of Embodiment of the Present Invention>
Hereinafter, details of embodiments of the present invention will be described with reference to the drawings. In addition, you may combine arbitrarily at least one part of embodiment described below.
A PDU (Protocol Data Unit) used in layer 2 communication is often referred to as a “frame”, and a PDU used in layer 3 communication is often referred to as a “packet”. However, in the following description, “packet” and “frame” "Is used in the meaning of PDU.

[Connection form of communication system]
FIG. 1 is a diagram showing a connection form of a communication system according to an embodiment of the present invention.
As shown in FIG. 1, the communication system of the present embodiment includes a communication network NW including a plurality of communication nodes N0 to N6, and a management device 4 that communicates with one communication node N0 via a higher level network 3. Yes.

In the following description, when common items of the communication nodes N0 to N6 are described, “N” is used as a representative code of the communication nodes N0 to N6.
In the communication network NW of FIG. 1, each communication node N is connected to one or a plurality of adjacent nodes via a transmission path 5 so as to be communicable. The communication medium constituting the transmission path 5 may be a transmission line such as a one-core or two-core optical fiber or a metal line, or may be a wireless communication path using a wireless LAN or the like.

In the communication network NW of FIG. 1, the communication node N0 is the master station 1, and the communication nodes N1 to N6 are the slave stations 2. The communication nodes N1 to N6 of the slave station 2 have a ring topology (N1 → N2 → N3 → N6 → N5 → N4 → N1).
Further, one communication node N6 in the ring is connected to the communication node N0 of the master station 1 by the transmission path 5. Therefore, for example, the path from the communication node N1 to the communication node N0 can be made redundant into two paths of N1->N2->N3->N6-> N0 and N1->N4->N5->N6-> N0.

The communication node N0 of the master station 1 is communicably connected to an upper network 3 such as the Internet. The terminal device 6 is connected to the communication nodes N1 to N6 of the slave station 2 so that they can communicate with each other.
In FIG. 1, for simplification of illustration, only one slave station 2 (communication node N3) is connected to the terminal device 6, but in practice, all the slave stations 2 (communication nodes N1 to N6) are connected. Are communicably connected to the terminal device 6.

The management device 4 includes a computer device capable of IP communication such as a server computer or a personal computer. The management device 4 can communicate with a predetermined terminal device 6 by IP communication via the upper network 3 and the master station 1 and communication within the communication network NW.
The terminal device 6 includes a computer device capable of IP communication such as a personal computer. The terminal device 6 can communicate with the management device 4 by communication within the communication network NW and IP communication via the master station 1 and the upper network 3.

[Usage form of communication system]
The usage forms of the communication network NW in FIG.
Use form 1: Remote control for a plurality of power distribution devices such as high-voltage switches managed by an electric power company, and remote sensing for the power distribution equipment Use form 2: Remote control for a plurality of manufacturing machines installed in a factory, and production thereof Remote sensing for machines Usage 3: Remote monitoring of in-house computer equipment and shared equipment connected to this equipment

In the usage mode 1, the terminal device 6 functions as a control command relay device for the management device 4 to remotely control the operation of a power distribution device (not shown).
For example, when controlling a power distribution device connected to the terminal device 6 of the communication node N1, the management device 4 transmits a downlink frame including a control command for the power distribution device to the terminal device 6 of the communication node N1. . The terminal device 6 controls the operation of the power distribution device according to the control command included in the received downlink frame.

In the usage mode 1, the terminal device 6 functions as a relay device for transmitting sensing information related to power distribution equipment to the management device 4.
For example, when collecting sensing information of a power distribution device connected to the terminal device 6 of the communication node N1, the terminal device 6 transmits an upstream frame including sensing information measured by various sensors (such as a voltmeter) of the power distribution device. It transmits to the management apparatus 4. The management device 4 totals the sensing information included in the received upstream frame.

Similarly, in the usage mode 2, the terminal device 6 functions as a control command relay device for the management device 4 to remotely control the operation of the manufacturing machine (not shown).
For example, when controlling a manufacturing machine connected to the terminal device 6 of the communication node N1, the management device 4 transmits a downlink frame including a control command for the manufacturing machine to the terminal device 6 of the communication node N1. . The terminal device 6 controls the operation of the manufacturing machine according to the control command included in the received downstream frame.

In the usage mode 2, the terminal device 6 functions as a relay device for transmitting sensing information related to the manufacturing machine to the management device 4.
For example, when collecting sensing information of a manufacturing machine connected to the terminal device 6 of the communication node N1, the terminal device 6 transmits an upstream frame including sensing information measured by various sensors (for example, a speedometer) of the manufacturing machine. It transmits to the management apparatus 4. The management device 4 totals the sensing information included in the received upstream frame.

In usage mode 3, the terminal device 6 is composed of a personal computer such as a desktop personal computer or a notebook personal computer used by employees in the company.
The shared device is communicably connected to a personal computer and includes an OA device such as a printer, a copier, or a facsimile that is used with the computer.

For example, paying attention to the terminal device 6 of the communication node N1, the terminal device 6 displays information on the operation contents when there is a terminal operation (for example, Internet access to an unauthorized homepage) that violates a predetermined in-house rule. The included upstream frame is transmitted to the management apparatus 4.
Further, when paying attention to the shared device of the terminal device 6 of the communication node N1, the terminal device 6 has the operation contents when the operation of the shared device violates a predetermined internal rule (for example, printing of a predetermined number of sheets or more). An uplink frame including information is transmitted to the management apparatus 4.

As described above, the communication network NW according to the present embodiment is used as a communication network for an administrator who operates the management apparatus 4 to remotely monitor target devices (such as power distribution equipment and manufacturing machines) corresponding to each communication node N. can do.
In this case, the communication node N is installed close to the target device, and the topology becomes complicated as the number of target devices increases. For this reason, the topology of the communication network NW is not limited to the simple ring shape shown in FIG. 1, but may be a complex topology such as a form including a plurality of rings or a tree shape or a mesh shape.

Of the communication frames relayed by the communication node N, the data included in the upstream frame addressed to the management apparatus 4 is referred to as “monitoring data”. The monitoring data includes “sensing information” relating to the above-described power distribution equipment and manufacturing machines, “information indicating operation contents” for personal computers and shared devices, and the like.
Further, in the communication system of FIG. 1, a connection form in which one communication node N is associated with a plurality of terminal devices 6, or a connection form in which a plurality of power distribution facilities or manufacturing machines are associated with one terminal device 6. It may be.

[Configuration of communication device]
FIG. 2 is a block diagram illustrating an internal configuration of the communication device 10. The communication device 10 of FIG. 2 can be employed for all or part of the communication nodes N of the communication network NW of FIG.
As illustrated in FIG. 2, the communication device 10 according to the present embodiment includes a control unit 11, a storage unit 12, a communication unit 13, a power supply unit 14, and a clock unit 15. The storage unit 12 includes a nonvolatile memory 16 and a volatile memory 17.

The control unit 11 includes a CPU (Central Processing Unit). The control unit 11 is connected to the communication unit 13 and the memories 16 and 17 via the internal bus 18. The control unit 11 comprehensively controls each unit of the communication node N.
The nonvolatile memory 16 includes, for example, an HDD (Hard Disk Drive), an SSD (Solid State Drive), a USB (Universal Serial Bus) memory, an SD memory card, and the like. The volatile memory 17 is composed of, for example, a DRAM (Dynamic Random Access Memory), an SRAM (Static Random Access Memory), or the like.

In the nonvolatile memory 16, a boot loader program, an OS (Operating System) program, an application program, and the like are installed. The application program includes a communication control program.
The volatile memory 17 stores a communication control program read from the nonvolatile memory 16 by the control unit 11 and is used as a working memory for the program.

The communication unit 13 is a communication interface having a communication function according to the path control protocol in the communication network NW and a communication function according to the communication protocol with the terminal device 6.
The control unit 11 executes communication control for the communication unit 13 in accordance with a communication control program stored in the volatile memory 17. The communication unit 13 transmits and receives IP packets (hereinafter simply referred to as “packets”) to and from adjacent nodes in accordance with instructions from the control unit 11.

For example, the control unit 11 performs packet transmission / reception with an adjacent node belonging to the communication network NW in accordance with a predetermined path control protocol (AODV, DYMO, etc.) adopted in the communication network NW. Specifically, the communication path is established and the packet is transferred using the established path.
In addition, the control unit 11 performs transmission / reception of a communication frame with the terminal device 6 according to a predetermined communication protocol adopted with the terminal device 6.

The power supply unit 14 includes an AC / DC converter therein. The power supply unit 14 converts an external power source composed of a commercial power source (for example, 100 V AC) into a predetermined DC voltage, and supplies the converted DC voltage to each unit in the communication node N.
The clock unit 15 includes a clock circuit that independently counts the local time. The clock unit 15 is operated by a standby power source (not shown) such as a coin-type lithium battery without receiving power supply from the power supply unit 14, and the local time is It is always output to the control unit 11.

[Communication node routing protocol]
The communication node N configuring the communication network NW relays the packet received from the adjacent node according to a predetermined route control protocol, and transfers the packet in the “bucket relay system” within the communication network NW.
Specifically, the communication network NW of the present embodiment employs a distance vector type routing protocol in which each communication node N searches for a route after a route request is generated.

Examples of the routing protocol include AODV (Adhoc ON-Demand Distance Vector) and DYMO (Dynamic MANET On-Demand). In this embodiment, AODV is assumed.
In AODV, all the communication nodes N (communication nodes N0 to N6 in FIG. 1) hold a “routing table (route table)” and use the routing table of their own node without including route information in the packet. Route control.

That is, AODV is an autonomous route control protocol, and is assumed to be a dynamic network in which the disappearance and appearance of the communication node N occur relatively frequently. Accordingly, the network path is not permanent, and a specific link is broken due to various failures, or a new communication node N appears and a new link appears.
In order to follow such a change in the network and each communication node autonomously establishes a route, a “route entry” having the following contents is defined for each end point (destination) in the routing table.

(Contents of route entry)
“Destination IP Address”: Destination IP Address
"Destination Sequence Number": Destination Sequence Number
“Destination Sequence Number Valid Flag”: Valid Destination Sequence Number flag
"Other flags": Other state and routing flags
(eg, valid, invalid, repairable, being repaired)

"Network Interface": Network Interface
“Hop Count”: Hop Count (number of hops needed to reach destination)
"Next Hop": Next Hop
"Precursor list": List of Precursors
“Validity Period”: Lifetime (expiration or deletion time of the route)

The “end point sequence number” is an index for indicating the newness of the route. A larger end point sequence number indicates a new route. Control message loops are prevented by not using the old sequence number path.
Further, when a path change such as a link disconnection occurs, the corresponding end point sequence number is incremented and diffused throughout the network by RERR. This notifies that a new change has occurred in the route of the network and prompts to invalidate the route.

The “valid period” represents a period during which the route entry is valid. The recommended value of the valid period is, for example, 3 seconds in AODV. The valid period is updated every time the route entry is used, and the route entry that is not used beyond the valid period from the last data communication becomes invalid.
Therefore, once the route entry becomes invalid because the valid period has been exceeded, the route entry cannot be used for packet relay unless the route entry is validated again by a new RREQ.

In AODV, the following five types are defined as route control messages. The communication node N exchanges these messages to maintain the route in an optimal state.
RREQ (Route Request) message RREP (Route Reply) message RERR (Route Error) message RREP-ACK (Route Reply Acknowledgment) message HELLO message

RREQ is a control message that is flooded when the source node requests a route search. RREP is a control message that the destination node transmits in unicast to the source node in response to RREQ. RERR is a control message that an intermediate node that has detected a link break notifies the source node.
RREP-ACK is a control message used for one-way link detection. HELLO is a control message for checking the survival status of adjacent nodes.

For example, in the communication network NW of FIG. 1, it is assumed that the communication node N3 makes a route request for route search with the communication node N5.
In this case, the source communication node N3 floods RREQ with the communication node N5 as the destination address. The communication nodes N2, N1, N4, and N6 that have received the RREQ also flood the same RREQ. As a result, the RREQ arrives at the communication node N5 through the clockwise and counterclockwise routes.

Each communication node N generates a reverse path (hereinafter referred to as “reverse path”) with respect to the transmission source address (IP address of the communication node N3) when transferring the RREQ.
The communication node N5 set as the destination returns a RREP addressed to the communication node N3 by unicast to the first RREQ (in the example of FIG. 1, RREQ relayed in the order of N3 → N6 → N5). The RREP is delivered to the communication node N3 that is the transmission source through the reverse path generated when the RREQ is transferred.

Each communication node N generates a forward path (hereinafter referred to as “forward path”) for the destination address when transferring the RREP. Thereby, a bidirectional path between the communication node N3 and the communication node N5 is established.
Thus, when the communication path between the communication node N3 and the communication node N5 is established, the communication node N3 starts transmitting data packets to the communication node N5.

[Configuration of communication device controller]
FIG. 3 is a block diagram showing an internal configuration of the control unit 11 of the communication apparatus 10 that can be employed as the communication node N of the communication network NW that conforms to AODV.
Assuming the above-described usage modes 1 to 3 regarding the communication network NW of the present embodiment, the more communication devices necessary for the communication network NW exist, the more terminal devices 6 are collection targets of sensing information and information indicating operation contents. The number of N increases.

Therefore, it is preferable that the communication device 10 employed in the communication node N can operate with low power consumption and can be manufactured at low cost in both hardware and software.
Therefore, in the communication apparatus 10 of the present embodiment, a control unit 11 including a single core CPU is employed, and for example, a real-time OS in which the control unit 11 executes a plurality of processes preemptively is employed. The control unit 11 executes an application program that can execute AODV communication control incorporated in the real-time OS.

As illustrated in FIG. 3, the control unit 11 includes a header determination unit 21, a protocol determination unit 22, an entry determination unit 23, and an AODV processing unit as functional units realized by executing an application program for AODV communication control. 24.
The AODV processing unit 24 is subdivided into a message processing unit 25, a packet relay unit 26, and a route search unit 27 as its internal functional units. Note that reference numeral 28 in FIG. 3 denotes a “search waiting queue” included in the volatile memory 17.

Packets received by the communication unit 13 or packets in the search waiting queue 28 are sequentially input to the header determination unit 21. The header determination unit 21 determines whether the packet is addressed to the own station or another station from the IP header of the input packet.
The header determination unit 21 sends a packet whose determination result is addressed to its own station to the protocol determination unit 22, and sends a packet whose determination result is addressed to another station to the entry determination unit 23.

The protocol determination unit 22 determines whether the packet is an AODV control message or another packet from the IP header of the input packet.
The protocol determination unit 22 sends a packet that is a control message having a determination result of AODV to the message processing unit 25 of the AODV processing unit 24. The protocol determination unit 22 sends a packet whose determination result is other (for example, a data packet addressed to the own station) to the general-purpose processing unit 29 that performs processing such as communication control with the management device 4 or the terminal device 6.

The entry determination unit 23 determines from the destination information in the IP header of the input packet whether the destination route entry is valid, invalid, or no entry.
The entry determination unit 23 sends a packet whose determination result is valid to the packet relay unit 26 of the AODV processing unit 24. The entry determination unit 23 sends a packet whose determination result is invalid or no entry to the route search unit 27 of the AODV processing unit 24.

The message processing unit 25 performs a predetermined process corresponding to the type of the control message of the packet on the input packet (AODV control message addressed to the own station).
For example, when the input control message is RREQ, the message processing unit 25 generates a reverse path for the transmission source. Further, the message processing unit 25 compares the RREQ information and the information held by itself with respect to the route information such as the sequence number and the hop count number, and updates the own route information when the RREQ information is newer. .

  If the input control message is RREP, the message processing unit 25 generates a forward path for the destination. Further, the message processing unit 25 compares the RREP information and the information held by itself with respect to the route information such as the sequence number and the hop count, and updates the route information when the RREP information is newer. .

The packet relay unit 26 performs packet relay processing on the input packet (a data packet addressed to another station and having a valid route entry or an AODV control message) based on the information of the corresponding route entry. .
Specifically, the relay processing performed by the packet relay unit 26 includes “packet transfer processing” and “validity period update processing”.

The packet forwarding process executed by the packet relay unit 26 controls the communication unit 13 to transmit the input packet addressed to the other station to the next-hop communication node N included in the currently valid route entry. It is processing to do.
The valid period update process executed by the packet relay unit 26 includes four paths (specifically, “forward path”, “reverse path”, “forward path The valid period of “next hop” and “next hop of reverse path”) is updated to a predetermined set value.

The route search unit 27 performs route search processing on the input packet (a data packet addressed to another station and the route entry is invalid or does not exist, or an AODV control message).
The route search process executed by the route search unit 27 includes a process of issuing an RREQ including a destination included in an input packet addressed to another station, a process of temporarily storing the issued RREQ in a search waiting queue, And processing for causing the communication unit 13 to transmit the issued RREQ.

[Conventional packet relay processing]
FIG. 4 is an explanatory diagram showing a conventional packet relay process.
As shown in FIG. 4, the conventional packet relay processing executed by the packet relay unit 26 includes a packet transfer task 31 and a period update task 32.
The packet transfer task 31 is a task corresponding to the above-mentioned “packet transfer process”, and the period update task 32 is a task corresponding to the above-described valid period update process.

In the conventional packet relay process, the packet relay unit 26 preemptively executes the packet transfer task 31 and the period update task 32.
In the conventional example of FIG. 4, when the packet transfer task 31 searches for a route entry from the IP header of a packet addressed to another station and determines a relay destination, the packet transfer task 31 validates the route entry before transmitting the packet to the relay destination. The period “update request” is transmitted to the period update task 32. The update request includes the source address and destination address extracted from the IP header of the received packet.

The period update task 32 that has received the update request updates the valid period of the route entry currently valid to the set value. The route entries for updating the validity period are a total of four entries: a route for the packet destination (forward route), a route for the packet transmission source (reverse route), and a route for each next hop related to the forward route and the reverse route.
The update of the valid period of these routes is executed based on the destination address and the source address of the IP packet notified by the update request.

Specifically, the period update task 32 searches for the route entry of the forward route corresponding to the destination of the IP packet, and updates the valid period of the route entry to the set value.
The period update task 32 searches for a route entry corresponding to the next hop included in the route entry of the forward route, and updates the validity period of the route entry to the set value.

Further, the period update task 32 searches for a route entry of a reverse route corresponding to the transmission source of the IP packet, and updates the valid period of the route entry to a set value.
The period update task 32 searches for a route entry corresponding to the next hop included in the route entry of the reverse route, and updates the valid period of the route entry to the set value.
When the update of the valid period by the period update task 32 is completed, the packet transfer task 31 starts transmitting a packet to the determined relay destination.

As described above, in the conventional relay process, the valid period of the valid route entry is updated, and then the packet addressed to the other station is transmitted.
The conventional relay process of FIG. 4 is implemented by setting the priority of the period update task 32 higher than the priority of the packet transfer task 31.

[Problems and solutions of conventional relay processing]
In the conventional packet relay process of FIG. 4, when a packet addressed to another station is received, the valid period of the route entry is updated prior to the transmission of the packet, and the packet is transmitted after this update is completed.
The reason is that according to the AODV communication standard, if a data packet is not relayed within the valid period of the route entry, the route entry becomes invalid and a new route search is required. This is thought to be due to the priority.

However, in conventional packet relay processing in which relay processing is performed in the order of packet reception → validity period update → packet transmission, the latency of packet transfer processing increases by the amount including the update of the validity period.
As a result, for example, the throughput of session-type data communication is lowered, and there is a problem that the communication speed of the entire network is deteriorated. This problem is significant in a communication network NW including a large number of communication nodes N as in the above-described usage modes 1 to 3.

If the main purpose is to secure the throughput of the communication network NW, as long as the route entry is valid, it is considered that it is sufficient that each communication node N quickly transfers the packet and then confirms the valid period of the route entry later.
Further, it is not always necessary to update the effective period of the route entry at the same timing as the packet transfer, and even if the start of the effective period update process is slightly delayed, it is considered that there is no particular problem.

In view of this, the present embodiment employs a configuration in which the route entry valid period update process is performed after the packet transfer is completed.
That is, in the communication device 10 according to the present embodiment, when the control unit 11 finds a corresponding route entry for a packet addressed to another station, the control unit 11 executes the packet transfer process in advance and uses the packet after completing the transfer process. Update the validity period of the route entry. As a result, in the packet relay process, the processing delay is omitted by the update time of the effective period, and the packet transfer latency is improved.

[Packet Relay Processing in First Embodiment]
FIG. 5 is an explanatory diagram illustrating packet relay processing according to the first embodiment.
As shown in FIG. 5, the packet relay process of the first embodiment executed by the packet relay unit 26 includes a packet transfer task 41 and a period update task 42.
The packet transfer task 41 is a task corresponding to the above-mentioned “packet transfer process”, and the period update task 42 is a task corresponding to the above-described valid period update process.

Also in the packet relay process of the first embodiment, the packet relay unit 26 preemptively executes the packet transfer task 41 and the period update task 42.
In the first embodiment of FIG. 5, when the packet transfer task 41 searches for a route entry from the IP header of a packet addressed to another station and determines a relay destination, the packet transfer task 41 performs packet transmission to the relay destination and performs this transmission. Immediately after completing the above, an “update request” of the validity period of the route entry is transmitted to the period update task 42. The update request includes the source address and destination address extracted from the IP header of the received packet.

The period update task 42 that has received the update request updates the valid period of the route entry currently valid to the set value. The route entries for updating the validity period are a total of four entries: a route for the packet destination (forward route), a route for the packet transmission source (reverse route), and a route for each next hop related to the forward route and the reverse route.
The update of the valid period of these routes is executed based on the destination address and the source address of the IP packet notified by the update request.

Specifically, the period update task 42 searches for the route entry of the forward route corresponding to the destination of the IP packet, and updates the valid period of the route entry to the set value.
The period update task 42 searches for a route entry corresponding to the next hop included in the route entry of the forward route, and updates the valid period of the route entry to the set value.

In addition, the period update task 42 searches for a route entry of a reverse route corresponding to the transmission source of the IP packet, and updates the valid period of the route entry to a set value.
The period update task 42 searches for a route entry corresponding to the next hop included in the route entry of the reverse route, and updates the valid period of the route entry to the set value.
When the update of the valid period by the period update task 42 is completed, the packet transfer task 41 can transfer a packet addressed to the next other station whose route entry is valid.

As described above, in the relay processing of the first embodiment, the valid period of the route entry is updated every time transfer of a packet addressed to another station is completed.
The relay process of the first embodiment of FIG. 5 is implemented by setting the priority of the packet transfer task 41 to be higher than the priority of the period update task 32.

As is clear from comparison between FIG. 4 and FIG. 5, according to the packet relay processing of the first embodiment, the latency from the packet reception to the packet transmission is increased by the time necessary for updating the valid period by the period update task 42. Can be reduced.
Therefore, by adopting the communication device 10 that performs such packet relay processing as the communication node N of the communication network NW, the packet transfer latency of each communication node N can be reduced, and the throughput of the communication network NW can be improved. .

  Further, in the packet relay processing (FIG. 5) of the first embodiment, the valid period is updated every time packet transfer completion is detected (update request is generated), so that the second embodiment (FIG. 6) described later is used. There is no need to provide a buffer 53 for storing a plurality of update requests. For this reason, it can be mounted more easily than in the case of a second embodiment described later (FIG. 6).

[Packet Relay Processing of Second Embodiment]
FIG. 6 is an explanatory diagram illustrating packet relay processing according to the second embodiment.
As shown in FIG. 6, the packet relay process of the second embodiment executed by the packet relay unit 26 includes a packet transfer task 51 and a period update task 52.
The packet transfer task 51 is a task corresponding to the above-described “packet transfer process”, and the period update task 52 is a task corresponding to the valid period update process.

Also in the packet relay process of the second embodiment, the packet relay unit 26 executes the packet transfer task 51 and the period update task 52 preemptively.
In the second embodiment of FIG. 6, when the packet transfer task 51 searches for a route entry from the IP header of the received packet addressed to another station and determines a relay destination, the packet transfer task 51 performs packet transmission to the relay destination and performs this transmission. Immediately after completing the above, an “update request” of the effective period of the route entry is transmitted to the period update task 52. The update request includes the source address and destination address extracted from the IP header of the received packet.

In the packet relay process (FIG. 5) of the first embodiment, the period update task 42 updates the valid period every time an update request is received. In the packet relay process (FIG. 6) of the second embodiment, the period update task 52 Stores the update requests immediately without processing them, and processes a plurality of update requests together.
That is, the period update task 52 temporarily stores the received update request in the update request buffer 53 provided in the volatile memory 17. Therefore, the CPU execution right is returned from the period update task 52, and the packet transfer task 51 can immediately shift to the next packet transfer.

After that, when a predetermined number (three in the example) of update requests are accumulated in the update request buffer 53, the period update task 52 updates the valid period of the route entry to the set value in order from the old update requests 1 to 3. .
The route entries for updating the validity period are a total of four entries: a route to the packet destination (forward route), a route to the packet transmission source (reverse route), and a next hop route related to the forward route and the reverse route.

The update of the valid period of these routes is executed based on the destination address and the source address of the IP packet notified by the update request.
The contents of the validity period update processing for each of the update requests 1 to 3 are the same as in the case of the first embodiment, and a description thereof will be omitted.

The period update task 52 can execute valid period update processing for the update requests 1 to 3 at a predetermined timing after packet transfer by the packet transfer task 51.
For example, it is preferable that the period update task 52 updates the effective period at a timing when the packet is not congested because the effective throughput can be improved. Note that the processing order may be executed from the oldest or may be executed collectively.

On the other hand, the following interruption process can be considered as a countermeasure when the valid period update process is not executed for a long time because the packet congestion state continues for a long time.
In other words, for example, the time of the oldest update request is regularly checked by an interrupt task set to high priority, and if the current elapsed time exceeds the allowable range, the update request is temporarily processed with high priority. You may decide to do it.

In the packet relay processing according to the second embodiment, the period update task 52 does not need to be set to a lower priority than the packet transfer task 51, but may be set to a lower priority.
When the period update task 52 is set to have a lower priority than the packet transfer task 51, the timing at which the packet transfer task 51 notifies the period update task 52 of the update request may be any timing in a series of packet relay processes. . When the period update task 52 has a low priority, the storage process of the update request in the buffer 53 is always executed after the packet transfer.

In the packet relay process of the second embodiment, the period update task 52 may record the time when the update request is received in the buffer 53 and correct the postponed time.
That is, when the time when the update request is received is T1, and the time when the update process is actually performed is T2, the time difference ΔT calculated by ΔT = T2−T1 is subtracted from the set value of the effective period (for example, 3 seconds). By setting the value to the valid period after update, the result is equivalent to the case of processing at the reception time.

As is clear from the comparison between FIG. 4 and FIG. 6, also in the packet relay processing of the second embodiment, the latency from the packet reception to the packet transmission is reduced by the time necessary for updating the valid period by the period update task 42. can do.
Therefore, by adopting the communication device 10 that performs such packet relay processing as the communication node N of the communication network NW, the packet transfer latency of each communication node N can be reduced, and the throughput of the communication network NW can be improved. .

In the packet relay process (FIG. 6) of the second embodiment, a plurality of update requests 1 to 3 and address information for each packet are accumulated in the buffer 53, and a plurality of packets are valid based on the accumulated address information. Period update processing is performed.
Accordingly, the execution period of the packet transfer process can be shortened as compared with the case of the first embodiment (FIG. 5) in which the valid period is updated every time the packet transfer is detected. For this reason, the throughput of the communication network NW can be further improved.

[Other variations]
The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

  For example, in the present embodiment, AODV and DYMO are exemplified as distance vector type routing protocols, but the present invention is not limited to this. Modifications such as improvements and modifications to a part of the protocol and combination with a link state type protocol are also possible.

  In addition, in the communication network NW of the present embodiment in which each communication node N performs route control according to a distance vector type routing protocol, the type of metric that is a route selection criterion includes the number of hops, the distance between nodes, and the like. A predetermined parameter may be selected depending on the protocol to be performed.

  Furthermore, in the communication network NW according to the present embodiment, the route selection algorithm applied to each communication node N is not limited to an algorithm that selects a route with an optimal metric (for example, the minimum number of hops), but also a suitable one (other Other algorithms may be used, such as selecting the second most suitable one in consideration of constraints.

DESCRIPTION OF SYMBOLS 1 Master station 2 Slave station 3 Host network 4 Management apparatus 5 Transmission path 6 Terminal apparatus 10 Communication apparatus 11 Control part 12 Storage part 13 Communication part 14 Power supply part 15 Clock part 16 Non-volatile memory 17 Volatile memory 18 Internal bus 21 Header Determination unit 22 Protocol determination unit 23 Entry determination unit 24 AODV processing unit 25 Message processing unit 26 Packet relay unit 27 Route search unit 28 Search wait queue 29 General-purpose processing unit 31 Packet transfer task 32 Period update task 41 Packet transfer task 42 Period update task 51 packet transfer task 52 period update task 53 update request buffer N0 to N6 communication node NW communication network

Claims (4)

A communication device used for a communication node that autonomously performs route control in a communication network,
A storage unit for storing a route entry for each destination node for which a valid period is set;
A control unit that performs relay processing for relaying a packet addressed to another station corresponding to the route entry during the effective period to an adjacent node,
The relay processing performed by the control unit includes
The packet transfer processing performed first;
And an update process that is performed after the transfer process and updates the valid period of the route entry corresponding to the packet to a set value. The update process includes
A process of accumulating address information of a plurality of transferred packets;
2. The communication device according to claim 1, further comprising: updating each of the valid periods of the route entries corresponding to the plurality of packets to set values based on the accumulated address information of the plurality of packets. The update process includes
Processing for detecting completion of transfer of the packet;
The communication apparatus according to claim 1, further comprising: a process of updating the valid period of the route entry corresponding to the packet to a set value every time completion of transfer of the packet is detected. A packet relay method performed by a communication device used for a communication node that autonomously performs route control in a communication network,
A first step of storing a route entry for each destination node for which a validity period is set;
A second step of performing a relay process of relaying a packet addressed to another station corresponding to the route entry during the effective period to an adjacent node;
The relay process in the second step includes
The packet transfer processing performed first;
A packet relay method including: an update process performed after the transfer process to update the valid period of the route entry corresponding to the packet to a set value.

Images