JP2006093885A - Communication method and node - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a communication method whereby packet transmission forms no loop in the case that networks with different control protocols are connected via a communication node terminated by a data link layer. <P>SOLUTION: The communication method is used for the communication node terminated by the data link layer provided between at least two difference networks. In this method , identification information is attached to a packet transmitted from the one network to the other network, whether or not the packet received from the other network includes identification information with prescribed contents is discriminated, and whether or not the packet is to be transferred to the one network is determined in response to a result of the discrimination. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention generally relates to the technical field of digital communication, and more particularly to a communication node and a communication method for connecting different networks.

  System design in this type of technical field uses the concept of network architecture to absorb differences in hardware, systems, etc., and to improve the efficiency of network construction. A network architecture or protocol stack includes multiple layers that are divided into various functions.

  FIG. 1 shows an outline of a protocol stack when two networks 1 and 2 are connected via a relay node. In the networks 1 and 2, various signal processes in the application layer, presentation layer, session layer, transport layer, Internet protocol layer (IP layer), data link layer, and physical layer are performed in order from the upper layer. For simplicity, the layer between the application layer and the IP layer is not shown in FIG. The relay node only needs to be able to relay packets between the networks 1 and 2. Therefore, it is not necessary for the relay node to support functions of all layers, and it is sufficient if functions below the IP layer or the data link layer are prepared.

  In data transmission from the network 1 to 2, data created in the application layer of the network 1 is processed along a route as indicated by a thick line in the figure and transmitted to the application layer of the network 2. In the network 1, a header is attached to data created in the application layer and transferred to a lower layer. In the lower layer, the content (header and data) received from the upper layer is used as the payload, and the payload with another header is transferred to the lower layer. Thereafter, the transfer contents are encapsulated in the same manner. Also in the IP layer, the contents received by the upper layer are set as the payload, and the header (IP header) attached is transferred to the data link layer. In the data link layer, a header (MAC header) is attached to the content received from the IP layer, which is transferred to the physical layer and transmitted to the relay node through a physical medium such as an electromagnetic wave, an electric signal, or an optical signal. The packet received at the physical layer of the relay node is transferred to the data link layer, the MAC header is removed, and the encapsulation is released. The contents after the release are transferred to the IP layer. In the IP layer, the transfer destination, transfer source, packet number, route information, and the like are analyzed from the contents of the IP header. According to this analysis result, a header is created, and the header and payload are transferred again to the data link layer. In the data link layer, a header (MAC header) is added to the content received from the IP layer, and the content is transferred to the physical layer and transmitted to the network 2. In the network 2, the received packet is decapsulated from the data link layer to the application layer, and finally, desired data reaches the application layer of the network 2.

  When the relay node is terminated at the IP layer, the packet can be transferred through an appropriate route based on the route information and the packet number included in the IP header. However, in this method, since the relay node needs to have all functions of the layers below the IP layer, the processing load on the relay node increases, and it becomes difficult to realize the relay node with a simple communication device. . There is also a concern that speeding up of relay operations such as packet routing and switching will be hindered.

  From the viewpoint of improving such inconvenience, termination of the relay node at the data link layer has also been studied. In the data link layer, since a routing route is not analyzed, a relay node between different networks can grasp only a packet transfer destination, a transfer source, and a payload. Even if information related to packet distribution is included in the MAC header, the contents of the MAC header are not maintained between different types of networks (may be maintained between the same types of networks). This is because a packet is transferred with a different MAC header attached to each network. For this reason, the relay node broadcasts the packet to the network connected thereto.

However, if the relay node broadcasts a packet without performing routing path management or the like, there is a concern that the packet is repeatedly transmitted and received between networks, and a so-called loop is formed. In order to avoid such a loop, the invention described in Non-Patent Document 1 uses a bridge that translates and converts protocols of both networks at the boundary between the networks. In the invention described in Non-Patent Document 2, functions of layers higher than the data link layer are supported by the relay node.
Rich Seifert (Author), Akira Mamiya (translation), "LAN switching thorough explanation", Nikkei BP, 6.6.1 Using SR-TB, p. 270-274 Rich Seifert (Author), Akira Mamiya (Translation), “LAN switching thorough explanation”, Nikkei BP, 6.6.2 Using router, p. 269-270

  However, in the method according to the invention described in Non-Patent Document 1, the relay node provided between the networks needs to understand and process both protocols of the connected networks, and the calculation burden is not small. Furthermore, it is necessary to prepare bridges according to various types of networks, which hinders easy network construction. As in the invention described in Non-Patent Document 2, if the function of the layer higher than the data link layer is supported, the packet link is terminated at the data link layer, which is contrary to the purpose of transferring the packet easily and at high speed. Become.

  The present invention has been made to address at least one of the above problems, and the task is terminated at the data link layer, which prevents packet transmission between networks with different control protocols from forming a loop. A communication node and a communication method are provided.

  In the present invention, a communication method in a communication node terminated at a data link layer provided between at least two different networks is used. This method adds identification information to a packet transmitted from one network to the other network, determines whether or not the packet received from the other network includes identification information of a predetermined content, and responds to the determination result. Thus, it is determined whether or not to transfer the packet to one network.

  According to the present invention, when a network having a different control protocol is connected via a communication node terminated at the data link layer, packet transmission can be prevented from forming a loop.

According to one aspect of the present invention, in a communication node terminated at a data link layer provided between two different networks having different control protocols, identification information is added to a packet transmitted from one network to the other, Whether or not to transfer the packet to one network is determined according to whether or not the packet received from the other network includes identification information having a predetermined content.
The boundary device between the networks checks whether the packet has been received in the past by checking the identification information, and determines whether or not the packet should be transferred to the network. Can be avoided. Since it is terminated not at the IP layer but at the data link layer, the relay node (communication node) does not require IP address management related to packet routing. The relay node does not need to understand both communication control protocols, but only needs to understand the communication control protocol of one network. It is not necessary to provide a bridge for converting the protocol at the relay node. Therefore, high-speed packet transfer by terminating at the data link layer can be realized.

  According to an aspect of the present invention, the identification information includes binary information corresponding to a transmission direction of a packet between networks. Thereby, the direction of packet transmission can be easily determined.

  According to an aspect of the present invention, the identification information includes a network identifier indicating the identity of the network. Since not only the direction of packet transmission but also the identity of the transmission source network can be determined, more advanced loop avoidance control can be performed.

  According to an aspect of the present invention, in a communication system having at least two different networks and at least first and second communication nodes terminated at a data link layer provided between the networks, the first and second The second communication node creates a network identifier indicating the identity of one network, broadcasts a broadcast packet including the network identifier to the one network, and is created by the first and second communication nodes. If the network identifiers are different, one of the network identifiers is given priority according to a predetermined criterion.

  As a result, the communication node can autonomously determine the network ID, and the necessity of comprehensively managing the network IDs of all networks can be eliminated.

  In one aspect of the present invention, the one network identifier is prioritized based on the size of the physical address values of the first and second communication nodes. Thereby, the priority can be determined without providing a new information field indicating the priority of the network ID in the packet.

  In one aspect of the present invention, any one communication node is determined based on a predetermined criterion, and communication nodes other than the determined communication node are prohibited from broadcasting a packet to the one network. As a result, it is possible to avoid the formation of a packet transmission loop without adding a field such as identification information to the packet separately.

  FIG. 2 shows an overall network configuration according to an embodiment of the present invention. FIG. 2 shows an internal network 200, three external networks 201, 202, and 203, and four boundary devices 211, 212, 213, and 214. The boundary devices 201 to 204 provided in the internal network 200 are equipped with a new function according to the present invention.

  The internal network 200 is a wireless LAN mesh network based on, for example, the IEEE 802.11 standard. The external networks 201 to 203 are, for example, Ethernet (registered trademark) wired LAN networks. These specific networks are examples, and the internal and external networks can be various networks. However, in the present invention, the internal network and the external network are different types of networks, that is, networks having different communication control protocols. Whether the network is internal or external depends on whether the boundary device 211 understands the communication control protocol. That is, the illustrated boundary devices 211 to 214 understand the communication control protocol of the internal network, but it is not essential to understand the control protocol of the external networks 201 to 203.

  The boundary devices 211 to 214 are also called gateway devices (GW1 to GW4), and relay packets transmitted between the internal and external networks by a method described later. The boundary devices 211 to 214 are terminated at the data link layer. In the illustrated example, the first boundary device 211 is connected to the first external network 201. The second boundary device 212 is connected to the first and second external networks 201 and 202. The third and fourth boundary devices 213 and 214 are connected to the third external network 203.

  FIG. 3 is a block diagram that focuses on the functions particularly related to the present invention among the various functions of the boundary device. The boundary devices 211 to 214 in FIG. 2 all have the same configuration and function, and the first boundary device 211 will be described as a representative thereof.

  The boundary device 211 includes first and second interfaces 302 and 304 and a packet transfer control unit 306. The packet transfer control unit 306 includes a transmission / reception interface determination unit 312, an identification information setting unit 314, a packet discard unit 316, and a packet transfer unit 318.

  The first interface 302 is connected to an internal or external network, and transmits / receives packets to / from the network. The second interface 304 is also connected to an internal or external network, and transmits / receives packets to / from the network. For example, the first interface 302 may be connected to the first external network 201, and the second interface 304 may be connected to the internal interface 200. The number of interfaces is prepared according to the number of networks connected to the boundary device. For example, for the second boundary device 212, three interfaces to the internal network and the first and second external networks are required. The interface may be physically provided individually, or a plurality of interfaces may be virtually provided by using a technique such as VLAN (Virtual LAN) regardless of the number of physical devices. In any case, two or more physical or virtual interfaces are provided so that communication with each network is appropriately performed.

  The packet transfer control unit 306 performs an operation of further transferring or discarding the packet received through the interface.

  The transmission / reception interface determination unit 312 analyzes the header of the received packet, and determines the transfer destination (destination) address, transfer source address, identification information, and the like of the packet. The identification information may be only binary information (0 or 1) corresponding to the transmission direction of the packet, and may further include a network identifier (network ID) of the network to which the transfer source device belongs. In this embodiment, the identification information includes a network ID in addition to the binary information.

  The identification information setting unit 314 specifically sets the content of the identification information. In this embodiment, 0 or 1 information is set as identification information according to the packet transfer direction. For example, identification information having a content of “1” is given to a packet transferred from the internal network 200 to the first external network 201. The information to be given may be 1-bit information as described above, but may be other-bit information from the viewpoint of determining the transmission direction of more packets. For example, the second boundary device 212 may use more than one bit number to distinguish packets from the internal network to the first external network and packets to the second external network. Good. Furthermore, the identification information setting unit 314 generates a network identifier (network ID) periodically, irregularly, or as necessary, and adds it to the identification information.

  The packet discard unit 316 discards the received packet based on the determination result of the transmission / reception interface determination unit 312 (particularly, the determination result of the content of the identification information) so that it is not further transferred.

  The packet transfer unit 318 transfers the received packet toward the destination indicated by the determination result of the transmission / reception interface determination unit 312. That is, the packet transfer unit 318 transfers the packet to the interface connected to the network to which the packet destination belongs.

  FIG. 4 is a flowchart showing the operation of the boundary device in one embodiment of the present invention.

  In step 402, the border device receives the packet.

  In step 404, it is determined through which interface 302, 304 the received packet has arrived. That is, the connection destination network of the receiving interface is determined. Thereby, it is determined at least whether the packet is received from the internal or external network. If the packet came from an external network, flow proceeds to step 406.

  In step 406, the packet transfer destination is determined. That is, the connection destination network of the transfer destination interface is determined. As a result, it is determined at least whether the packet is going to be transferred to the internal or external network. If the transfer destination is the internal network 200, the flow proceeds to step 408.

  In step 408, the packet identification information is set to a predetermined content, and the packet is broadcast in the internal network 200. The predetermined content includes a direction flag indicating the direction in which the packet is transferred, and a network ID indicating the identity of the transfer source network. As an example, the direction flag is set to “1” for the direction from the external network to the internal network, otherwise “0” is set. As described above, the direction flag may be expressed by a number of bits larger than 1 in order to distinguish more directions.

  Step 410 is performed when it is determined in step 406 that the transfer destination is an external network. In this step, it is determined based on the network ID in the identification information whether or not the transfer destination external network is the same as the transfer source external network. The external network ID of the transfer source is included in the identification information in the packet, but the external network ID of the transfer destination is determined when the interfaces 302 and 304 of the boundary device are connected to the external network. Alternatively, the network ID may be determined at another timing, but in any case, the network ID needs to be determined before step 410 is performed. If the transfer destination external network is the same as the transfer source external network, the flow proceeds to step 412.

  In step 412, the packet is discarded and the packet is not forwarded to the external network indicated as the forwarding destination. Thereby, it can be avoided that the same packet is transferred to the external network to form a loop.

  Step 414 is performed when it is determined in step 410 that the transfer destination external network is not the same as the transfer source external network. In this step, the packet is transferred to the external network indicated as the transfer destination.

  Step 420 is performed when the received packet comes from the internal network in step 404. In this step, the packet transfer destination is determined. That is, the connection destination network of the transfer destination interface is determined. If the transfer destination is the internal network 200, the flow proceeds to step 422.

  In step 422, the packet is broadcast in the internal network 200.

  Step 424 is performed when it is determined in step 420 that the transfer destination is an external network. In this step, the packet identification information is confirmed. If the direction flag in the identification information has a predetermined content such as “0”, the flow proceeds to step 426.

  In step 426, the packet is transferred to the external network indicated as the transfer destination. Since the direction flag of the identification information is “0”, the packet may not have been transferred from the external network to the internal network in the past, but may be newly transferred from the internal network to the external network. I understand.

  Step 428 is performed when the direction flag in the identification information does not have the predetermined content in Step 424 (for example, when the directionality flag is “1”). In this step, a packet transfer destination is further determined. That is, the relationship between the network to which the receiving interface is connected and the network indicated by the identifier is confirmed. In this step, it is determined based on the network ID in the identification information whether or not the transfer destination external network is the same as the transfer source external network. If the destination external network is the same as the source external network, the flow proceeds to step 430.

  In step 430, the packet is discarded and the packet is not forwarded to the external network indicated as the forwarding destination. Thereby, it can be avoided that the same packet is transferred to the external network to form a loop.

  Step 432 is performed when it is determined in step 428 that the transfer destination external network is not the same as the transfer source external network. In this step, the packet is transferred to the external network indicated as the transfer destination.

  The network ID that distinguishes each of the external networks should be uniquely determined. When one external network (for example, 201) is connected to a plurality of boundary devices (for example, 211, 212), the external network should be specified by the same network ID in each boundary device. From such a viewpoint, a control unit (not shown) that collectively manages network IDs of a plurality of external networks may assign appropriate network IDs to the respective external networks. However, the management burden of the network ID of such a general control unit is not necessarily small. In such a case, it is desirable that each boundary device autonomously determines the network ID. In the following second embodiment, a technique for setting an appropriate network ID in the boundary device will be described.

  FIG. 5 shows a flowchart of a method for setting an appropriate network ID. This operation may be performed not only when the boundary device is connected to the external network, but also periodically or irregularly (as necessary). For convenience of explanation, this flow is mainly described as being performed by the boundary devices 211 and 212 in FIG. 2, but similar processing is performed in other boundary devices. The flow begins at step 502 and proceeds to step 504.

  In step 504, the first boundary device 211 determines a network ID (initial network ID) NW1 for the first external network 201. This determination is provisional, and another network ID may be finally adopted depending on the contents of subsequent processing. The contents of the network ID can be set using any existing method, and may be set randomly, generated according to some calculation method, or selected from some list. However, it may be set manually.

  In the current example, since the second boundary device 212 is also connected to the first external network 201, the second boundary device 212 also determines the network ID (NW2) by a similar method. The first and second boundary devices 211 and 212 determine network IDs independently of each other.

  In step 506, a broadcast packet including the initial network ID is created. The broadcast packet is broadcast regularly or irregularly to the external network corresponding to the initial network ID. In the current example, the first boundary device 211 broadcasts a notification packet including the network ID (NW1) to the first external network 211. Similarly, the second boundary device 212 broadcasts a notification packet including the network ID (NW2) to the first external network 211.

  In step 508, the boundary device determines whether or not a notification packet from another boundary device (another node) has been received. Since the first external network 201 is also connected to the second boundary device 212, the notification packet broadcast from the first boundary device 211 is not only the communication node in the first external network 201, The second boundary device 212 is also reached. Further, the notification packet broadcast from the second boundary device 212 reaches not only the communication node in the external network 201 but also the first boundary device 212. When the border device receives a notification packet from another border device, the flow proceeds to step 512. If the border device does not receive a broadcast packet from another border device, the flow proceeds to step 510. In the latter case, the second boundary device 212 broadcasts a notification packet to the second external network 202, that is, the external network is connected to only one boundary device.

  In step 510, the initial network ID is determined as the formal network ID. For example, the network ID tentatively determined for the second external network 202 by the second boundary device 212 is formally determined as the network ID of the second external network 202 and used for subsequent communication. The Thereafter, the flow proceeds to step 516 and ends.

  Step 512 is performed when the boundary device receives a notification packet from another boundary device in Step 508. In the current example, the first boundary device 211 receives the notification packet from the second boundary device 212. As a result, the first boundary device 211 knows that some boundary device other than itself is connected to the first external network 201. In this step, the priority of the own and other nodes is determined. That is, the first boundary device selects either the network ID (NW1) determined by itself or the network ID (NW2) determined by another boundary node based on a predetermined determination criterion. Various methods can be used as to which network ID is prioritized. For example, the priority order of the boundary devices may be given and the network ID assigned by the high priority boundary device may be adopted. The priority order of the boundary devices may be determined based on the size of the medium access control (MAC) address (physical address). A field indicating priority may be provided in the broadcast packet. When the priority is determined using the MAC address, it is advantageous in that it is not necessary to provide such a special field. If the network ID (NW1) assigned by the first boundary device is prioritized, the flow proceeds to step 510 and the described processing is performed. Note that, when broadcast packets (network IDs) are received from more than two border devices, the highest priority among them is given priority.

  In the current example, the same processing is performed in the second boundary device 212. The second boundary device 212 selects either the network ID (NW2) determined by itself or the network ID (NW1) determined by another boundary device based on a predetermined determination criterion. Therefore, the priority criterion needs to be common to each boundary device.

  In step 514, the network ID determined by the other boundary device is determined as the network ID indicating the external network. Thereafter, the flow proceeds to step 516 and ends.

  In the present embodiment, any one of a plurality of network IDs indicating the same external network is selected based on a predetermined determination criterion. However, another network ID may be further created based on those network IDs. In short, it is only necessary to use one common network ID among a plurality of boundary devices connected to one external network.

  In the embodiment described below, when a plurality of boundary devices are connected to one external network, a packet transmission loop is avoided by limiting the number of boundary devices that can broadcast packets to the external device to one. Is done. FIG. 6 is a flowchart showing the operation of the boundary device according to the present embodiment. The flow begins at step 602 and proceeds to step 604.

  In step 604, the border device transmits a notification packet.

  In step 606, the boundary device determines whether a notification packet from another boundary device has been received.

  Step 608 is performed when a certain boundary device does not receive a notification packet from another boundary device in Step 606. In the situation leading to this step, no border device other than “a border device” is connected to the external network. Therefore, a “certain boundary device” is selected as a boundary device that can transfer a packet to an external network. Then, the packet transfer function of the boundary device to the external network is made to function effectively.

  Step 610 is performed when a boundary device receives a notification packet from another boundary device in Step 606. In the situation leading to this step, boundary devices other than “a certain boundary device” are connected to the external network. Accordingly, any one of the plurality of boundary devices connected to one external network is selected. As the selection criterion, the method related to the priority of the second embodiment can be used. For the selected boundary device, the process of step 608 is performed, and the flow proceeds to step 614 and ends. For the boundary device not selected, in step 612, the packet transfer function of the boundary device to the external network is disabled. Thereafter, the flow proceeds to step 614 and ends.

It is a figure which shows a network architecture. 1 shows an overall view of a network configuration according to an embodiment of the present invention. FIG. A block diagram relating to the main functions of the border device is shown. 3 is a flowchart illustrating an operation according to an exemplary embodiment of the present invention. 3 is a flowchart illustrating an operation according to an exemplary embodiment of the present invention. 3 is a flowchart illustrating an operation according to an exemplary embodiment of the present invention.

Explanation of symbols

200 Internal network; 201, 202, 203 External network; 211, 212, 213, 214 Border device;
302, 304 interfaces; 306 packet transfer control unit; 312 transmission / reception interface determination unit; 314 identification information setting unit; 316 packet discard unit; 318 packet transfer unit

Claims (9)

In a communication node terminated at a data link layer provided between at least two different networks,
Add identification information to packets transmitted from one network to the other,
Determine whether a packet received from the other network includes identification information of a predetermined content,
Whether to transfer the packet to one of the networks is determined according to the determination result. The communication method according to claim 1, wherein the identification information includes binary information corresponding to a transmission direction of a packet between networks. The communication method according to claim 1, wherein the identification information includes a network identifier indicating an identity of the network. In a communication system comprising at least two different networks and at least first and second communication nodes terminated at a data link layer provided between the networks,
The first and second communication nodes each create a network identifier indicating the identity of one network, and broadcast a broadcast packet including the network identifier to the one network,
When the network identifiers created by the first and second communication nodes are different, any one of the network identifiers is given priority according to a predetermined criterion. 5. The communication method according to claim 4, wherein the one network identifier is prioritized based on the size of the physical address value of the first and second communication nodes. One of the communication nodes is determined based on a predetermined criterion,
The communication method according to claim 4, wherein communication nodes other than the determined communication node are prohibited from broadcasting a packet to the one network. A communication node terminated at a data link layer provided between at least two different networks,
A determination means for determining a transfer destination and a transmission source of the received packet;
Means for adding identification information to a packet transmitted from one network to the other network, and the determination means determines whether or not the packet received from the other network includes identification information of a predetermined content, Whether to transfer the packet to one of the networks is determined according to the determination result. The communication node according to claim 7, wherein the identification information includes binary information corresponding to a transmission direction of a packet between networks. The communication node according to claim 7, wherein the identification information includes a network identifier indicating an identity of the network.

Images