JP2002185504A - Packet transfer repeating system and method for transfer control - Google Patents

Abstract

PROBLEM TO BE SOLVED: To rapidly process without increasing the number of entries of a retrieving route table even in a large-scale network. SOLUTION: A transfer repeating system in the network sorts the input packets by a sorter 12, adds to the packet information regarding coordinates of an address unit of the packet obtained by inquiring an address mapping unit for holding correspondence of the packet free from adding a coordinate label, to the coordinate section of a topology of the network. When the packet is sent to the adjacent packet transfer repeating system, a coordinate transfer processing unit 19 of a packet transfer repeating system sends the packet to one of the packet transfer repeating system, in which a distance in a multi- dimensional coordinate space of the packet transfer repeating system of a transmitting destination based on the respective coordinates of the adjacent one or more packet transfer repeating systems and the coordinates of the repeating system at the transfer destination.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention can reduce the number of entries in a coordinate search table to be searched regardless of the size of a network when deciding a transfer destination of packet transfer, thereby enabling high-speed search processing. And a transfer control method.

[0002]

2. Description of the Related Art A packet transfer network in which packet transfer repeaters such as routers are connected to each other has been known, and a packet transfer network using the Internet Protocol (IP) has been widely used. In a network using IP, addresses are assigned to devices such as routers and hosts in the network. The packet transfer source device writes the address of the destination device in the header of the packet and inputs the address into the network. The packet transfer relay device searches the routing table or the like stored in the packet transfer relay device for the destination address described in the packet, determines the next transfer destination packet transfer relay device or the connection line to be transmitted, and performs the transfer processing. I do.

In a conventional packet transfer relay device, as shown in FIG. 2, when a packet is input to an input port 11, the distribution unit 12 separates the packet from the route control protocol packet. The processing is performed by the routing control protocol processing unit 13. The path control protocol processing unit 13 is an OSPF (Open Shortest Path Firs
t) or processing of a dynamic routing protocol such as RIP (Routing Information Protocol) to create a routing table 15 used for packet transfer based on the Dijkstra algorithm or the Bellman Ford algorithm. Packets other than the route control protocol packet are sent to the normal transfer processing unit 14. Normal transfer processing unit 1
In step 4, the routing table 15 is searched for the destination address included in the packet, and the next transfer destination is selected and transferred.

FIG. 3 is a diagram showing a routing table of a conventional packet transfer relay device. 2 shows a routing table of a router installed in a node D in a case where transfer relay is performed using a conventional router in the packet transfer network shown in FIG. In FIG. 3, the destinations A to S are not aggregated (aggregated), and the destinations A to S are actually address prefixes, and the number of entries in the routing table 15 may be smaller due to the aggregation. Generally, the number of entries in the routing table 15 of each packet transfer relay device when a conventional router or the like is used increases according to the scale of the network. The routing table 15 is referred to when all packets are transferred, and the search time is approximately proportional to the logarithm of the number of entries.

[0005]

As described above, conventionally, in a packet transfer network such as the Internet, the addresses of the devices are usually assigned irrespective of the topology configuration of the network. In the packet transfer relay device, since an entry is created for each destination prefix in the routing table in the device, the routing table increases as the network scale increases, and it takes time to search the routing table. As the number of entries in the routing table increases, the amount of processing required for searching the routing table also increases,
It was necessary to increase the processing capacity of the apparatus.

In particular, a packet transfer network in which a number of wireless relay stations which are being realized or are expected to be realized in the future are arranged, for example, a wireless LAN network using IEEE 802.11, a next-generation mobile phone network, or Bluetooth is used. In an ad hoc network, it is expected that the scale of the network will be larger and / or the number of connection lines between packet transfer relay devices in the network will be larger than before. In such a network, the conventional method is expected to increase the amount of routing table search processing.

Accordingly, an object of the present invention is to solve these conventional problems and to provide a packet transfer relay device and a transfer control device capable of processing at high speed without increasing the number of entries in a routing table searched even in a large-scale network. It is to provide a method.

[0008]

In order to achieve the above object, a packet transfer control method according to the present invention provides a method for controlling each packet transfer repeater in multi-dimensional coordinates based on the mutual connection status of the packet transfer repeaters in a network. When a packet is transmitted from an arbitrary packet transfer relay device corresponding to a point in space, information on coordinates in a multidimensional coordinate space corresponding to the packet transfer relay device of the transmission destination is added to the packet. When a packet is transmitted to an adjacent packet transfer relay device, the number of destination packet transfer relay devices is determined based on the coordinates of one or more adjacent packet transfer relay devices and the coordinates of the source packet transfer relay device. The packet is transmitted to one of the packet transfer repeaters having the smallest distance in the dimensional coordinate space.

[0009] Each packet relay device in the network is associated with a point in the multi-dimensional coordinate space by using a dynamic routing protocol or by connecting each packet transfer relay device obtained by setting information. A topology mapping unit for causing each packet transfer relay device to correspond to a point in a multidimensional coordinate space based on information, a map storage unit for storing the correspondence generated by the topology mapping unit, and a packet transfer relay device in a network The map storage unit is searched for an inquiry including the address of the address, and is performed by an address map device having an inquiry processing unit that answers the coordinates of the node obtained based on the correspondence.

Each packet transfer relay device in the network makes an inquiry to the address map device as appropriate, and holds a correspondence table indicating the corresponding coordinates of the packet transfer relay device in the network.
In transferring the packet, the transfer source device of the packet or the packet transfer relay device directly accommodating the transfer source device accommodates the transfer destination device or the transfer destination device corresponding to the destination address with reference to the correspondence table. If the information relating to the coordinates of the packet transfer relay device (hereinafter, destination device) exists in the above correspondence table, if it does not exist, use the information obtained by making an inquiry to the address map device. Label information on the coordinates of the destination device is added to the packet.

In addition, the packet transfer relay device that has received the packet to which the label information has been added compares the coordinates of the destination device indicated by the label information with its own coordinates, and determines the next transfer destination transfer relay device and / or Outgoing connection line (hereinafter,
Next transfer destination) is determined and transfer is performed.

[0012] The packet transfer relay device monitors a fault condition of an adjacent packet transfer relay device and / or link, and when a fault occurs or recovers, the coordinates are close to each other in the multidimensional coordinate space. The failure notification may be sent to the packet transfer relay device. In that case, the packet transfer relay device that has received the notification may use the fault information to bypass the fault location when determining the next transfer destination based on the label information.

In a situation where the degree of a fault exceeds a threshold, such as the number of faults in a network or a stop period, the status is determined by using a dynamic routing protocol or by notification from a packet transfer relay device near the fault. Is recognized by the address map device, and the address map device in the multidimensional coordinate space of each packet transfer relay device based on the physical connection information of the connection line connecting each packet transfer relay device at that time. Re-association with points may be performed.

[0014]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 4 is a block diagram showing the concept of the address map device according to the present invention. The address map device according to the present invention is a device for associating each packet transfer relay device in the network with a point on the multidimensional coordinate space, and as shown in FIG. Topology mapping unit 22
, A map storage unit 23, and an inquiry processing unit 21. The address map device may physically be one obtained by adding each processing unit shown in FIG. 4 to one of the packet transfer relay devices. Therefore, it is needless to say that the address map device can be included in each packet transfer relay device, but it is also possible to install an independent address map device for each of a plurality of packet transfer relay devices. Alternatively, by including it in one of the plurality of packet transfer relay devices, it is possible to respond to requests from other packet transfer relay devices. The routing control protocol processing unit 13
OSPF used in P communication, PNNI (Private Network-Network Interface) used in ATM network
e) and the like, and has a function of reproducing and maintaining the physical topology of the network as shown in FIG. 1, for example.

In FIG. 1, white circles represent packet transfer repeaters, alphabets from A to S in the white circles represent addresses of the packet transfer repeaters, and straight lines connecting white circles represent connection lines between the packet transfer repeaters. ing.
The topology mapping unit 22 has a function of making each packet transfer relay device correspond to a point in the multidimensional coordinate space based on the physical topology of the network reproduced and held by the routing control protocol processing unit 13.

For an arbitrary graph (linear graph),
Algorithms for determining the flatness are known (for example, KSBooth and GSLueker, “Testing for the c
onsecutive ones property, interval graphs, and grap
h planarity using PQ-treeAlgorithm, "Jounal of Comp
uter and System Sciences, 13: 335-379, 1976). Also,
Using Repeat these algorithms, in the subgraph of non-planar graph, it is possible to identify the class to-ski graph K ₅ or K _3.3 in phase ones, based on any one of the edges included in the sub-graph (edge) By repeatedly performing the operation of removing the graph from the above graph, a planar graph which is a subgraph of the original non-planar graph can be obtained.
In addition, N. Nishizaki and N. Chiba. “Planar G
raphs: Thery and Algorithms. "Annals of Discrete Mat
hematics (32), North-Holland Mathematics Studies, 198
8 are known. It is known that the amount of calculation of the above-described algorithm is proportional to the size of the graph (the number of nodes and the number of edges). By using these algorithms in the topology mapping unit 22,
It is possible to associate the physical topology of the network on a multidimensional coordinate space.

FIG. 5 is a diagram showing an example of association of the network in FIG. 1 with the coordinate space. For example, there can be a plurality of correspondences to the multidimensional coordinate space by the network topology mapping unit 22 shown in FIG. 1, and FIG. 5 shows an example of the correspondence to the two-dimensional lattice coordinate space among them. is there. In the association in FIG. 5, each connection line between the packet transfer relay apparatuses is associated with a vector in which the size of each dimensional component in the multidimensional coordinate space is 1 or 0. In FIG. 5, white circles with A to S correspond to the respective packet transfer relay devices in FIG. 1, and (0, 1) and the like are the coordinates of each packet transfer relay device in association. The thick straight line between the white circles corresponds to the connection line between the packet transfer repeaters.

The correspondence to the multidimensional coordinate space created by the topology mapping unit 22 is stored in the map storage unit 23.
Is stored in The query processing unit 21 receives a query packet sent from another device such as a packet transfer relay device in the network, and stores the address of the packet transfer relay device in the network included in the query packet in the map storage unit 23. Referring to the correspondence of the network topology to the multidimensional coordinate space, an inquiry response packet including the coordinates of the node is created and sent to the above-mentioned device which is the source of the inquiry packet.

FIG. 6 is a block diagram of a packet transfer repeater according to an embodiment of the present invention. The packet transfer relay device of the present invention has a coordinate label creating unit 20, a local topology processing unit 17 And a coordinate transfer table 18 and a coordinate transfer processing unit 19 in its configuration. The packet arriving at the packet transfer relay device in FIG. 6 is first classified by the distribution unit 12 into a packet without a coordinate label and a packet with a coordinate label. Among the packets to which the coordinate label is not added, the routing control protocol packet is processed by the routing control protocol processing unit 13. Packets to which no coordinate label is added and which are not route control protocol packets are processed by the coordinate label creating unit 20. A packet to which a coordinate label is added and which is not a local topology packet described later is sent to the coordinate transfer processing unit 19.
Is processed. Further, a packet to which a coordinate label is added and which is a local topology packet is processed by the local topology processing unit 17.

When receiving a packet without a coordinate label (hereinafter, a data packet), the coordinate label creating unit 20 determines a destination address of the data packet with respect to the destination address of the data packet.
Create and send an inquiry packet to the address map device. The data packet is buffered while waiting for an inquiry response packet from the address map device. When the inquiry response packet from the address / map device arrives within the specified time, the coordinate label creating unit 20 extracts the coordinates from the inquiry response packet, adds the coordinates to the data packet as a coordinate label, and sends the data packet to the coordinate transfer processing unit 19. If the inquiry reply from the address map device does not arrive within the specified time, the routing table 15 is searched for the destination address of the data packet, and the next transfer device is determined and transferred, as in the case of a normal router. . For the packet with the coordinate label, the coordinate transfer processing unit 19 compares the coordinates of the destination device indicated by the coordinate label with its own coordinates, determines the next transfer destination, and transfers the packet.

FIG. 7 is a conceptual diagram of transfer by coordinates in the present invention. FIG. 7 shows a general image of transfer based on coordinates. Here, A is its own device, a white circle around A is a packet transfer relay device directly connected to A, and the packet transfer relay device to be the next transfer destination is determined by the direction from A to the destination device. If B is the destination device, the next transfer destination is C.

FIG. 8 is a diagram of an embodiment (1) of a transfer process using coordinates in the packet transfer relay device according to the present invention. The operation of the coordinate transfer processing unit of the packet transfer relay device D will be described with reference to the coordinate space shown in FIG.
In the coordinate transfer table, the directions and addresses of the adjacent packet transfer relay devices are described. If the destination device of the data packet is O,
The coordinate transfer processing unit 19 firstly determines the coordinates (5, 3) of the destination device.
Is subtracted from the coordinates (1, 2) as a vector to obtain a vector (4, 1) indicating the direction of the destination. An extended Hamming distance, which is the sum of the difference between each coordinate component between the vector indicating the direction of the destination and the direction of each entry in the coordinate transfer table, is calculated. Select as transfer destination.

FIG. 9 is a diagram of an embodiment (2) of a transfer process using coordinates in the packet transfer relay device according to the present invention. The operation of the coordinate transfer process of the packet transfer relay device D will be described based on the correspondence to the coordinate space shown in FIG. The address of the relay device of the next transfer destination is described in the coordinate transfer table in eight directions around the two-dimensional coordinate space. In the present embodiment (2), among the eight directions, for the direction in which there is no directly connected packet transfer relay device, the packet transfer relay device directly connected in the direction closest to the direction (if there are a plurality of them, the half-clock A coordinate transfer table is configured with the next transfer destination as the next transfer destination. If the destination device of the data packet is O, the coordinate transfer processing unit 19
First, the coordinates (1, 3) of the destination device are converted to the coordinates (1,
2) is subtracted as a vector to obtain a vector (4, 1) indicating the direction of the destination. Next, among the coordinate components of the vector indicating the direction of the destination, the vector is divided by the one having the largest absolute value (4), normalized by dividing the vector (1, 0.25), and further obtained by rounding up each component. With reference to the entry of the coordinate transfer table that matches (1, 1), F is selected as the next transfer destination.

As can be seen from the configuration of the coordinate transfer table in FIGS. 8 and 9, in the present invention, the number of entries in the coordinate transfer table is the size of the network, that is, the packet transfer relay device, connection line, destination included in the network. Since it is irrelevant to the number of devices, it is possible to cope with a large-scale network. The number of entries in the coordinate transfer table is clearer in FIG. 8 and FIG. 9 as is clear when comparing the coordinate transfer table shown in FIG. 8 and FIG. 9 with the routing table 15 of the conventional packet transfer relay device shown in FIG. Since it is smaller than FIG. 3, high-speed search is possible. Each search of the routing table 15 in the conventional system and the coordinate transfer table in the system of the present invention is a process performed for all data packets. Therefore, if the search process is performed at high speed, the entire packet transfer process will be speeded up. Will be.

Each packet transfer relay device transmits the coordinates of its own device, the coordinates of another transfer relay device adjacent to the device itself, and information on the presence or absence of a failure in another adjacent packet transfer relay device and / or connection line. Can be exchanged with other transfer devices. The packet used for this exchange is called a local topology packet. Specifically, the local topology packet includes, for example, a sequence of the coordinates of the own device and the coordinates of a device that is adjacent to the own device and can communicate with the current device (or a new communication is possible after the last time when the local topology packet was issued). Alternatively, a column of the coordinates of the device that has become impossible and the communication state) are described. Alternatively, the local topology packet may describe a sequence of vectors obtained by subtracting the coordinates of the own device, the coordinates of a device adjacent to the own device and currently communicable, and the own device coordinates as a vector. The local topology packet arriving at the packet transfer relay device from another packet transfer relay device is transmitted to the distribution unit 12 described above.
Is sent to the local topology processing unit 17 by the distribution according to. The local topology processing unit 17 updates the contents of the coordinate transfer table based on the information described in the local topology packet. Since the coordinate transfer table is a table for each direction from the own device and does not represent the topology of the entire network, the local topology packet may be delivered only to the vicinity of the own device in the multidimensional coordinate space.

FIG. 10 is a diagram showing an example (1) of the information distribution range of the local topology information according to the present invention. FIG. 10 shows an example of a limitation on the distribution range of the local topology packet related to the failure of the transfer device at the coordinates (p, q) indicated by the black circles. A circle indicates a transfer relay device, and a straight line connecting the circles indicates a connection line between the packet transfer relay devices. The devices around the faulty device send the local topology packets only to the transfer devices adjacent to each other, so that the distribution range is limited to the gray circle packet transfer relay device within two hops from the faulty device.

FIG. 11 is a diagram showing an example (2) of the information distribution range of the local topology information according to the present invention. FIG. 11 shows an example of the restriction on the distribution range of the local topology packet related to the failure of the transfer device at the coordinates (p, q) indicated by the black circles. A circle indicates a packet transfer relay device, and a straight line connecting the circles indicates a connection line between the packet transfer relay devices. By notifying fault information only in a direction parallel to the x and y coordinate axes from a device around the faulty device, the distribution range is limited to the gray-circular packet transfer relay device.

FIG. 12 is a diagram showing an example of the distribution range of the local topology information in FIG. FIG. 12 shows a transmission range when the transmission of the local topology packet is limited to one hop in one correspondence to the coordinate space shown in FIG. That is, the local topology packet from G is distributed to F, H, R, and Q, and the local topology packet from N is distributed to L, M, S, and O, respectively.

FIG. 13 is a diagram showing an example of a connection line failure in FIG. 5, and FIG. 14 is a diagram showing an example (1) of a packet transfer relay process in the packet transfer relay device in FIG.
In FIG. 14, when the coordinate transfer processing unit 19 performs the operation shown in FIG. 8, when the failure occurs in the connection line between the packet transfer processing D and F as shown in FIG. 2 shows the operation of the coordinate transfer processing unit 19 in the packet transfer relay device D for the packet No. That is, due to the occurrence of the failure, the entry of the direction (1, 1) is deleted from the coordinate transfer table of D. The vector of the destination direction is calculated, and as a result of the calculation of the extended Hamming distance, among the remaining entries, E having the minimum extended Hamming distance is selected as the next transfer destination.

FIG. 15 is a diagram showing an example (2) of the packet transfer relay process in the packet transfer relay device D in FIG. In FIG. 15, when the coordinate transfer processing unit 19 performs the operation shown in FIG. 9, when the failure occurs in the connection line between the coordinate transfer relay devices D and F as shown in FIG. 2 shows the operation of the coordinate transfer processing unit 19 in the packet transfer relay device D for the packet to the destination. That is, due to the occurrence of a failure, the coordinate transfer table of D
Is changed in accordance with the rule described with reference to FIG. Calculate the destination direction vector,
As a result of normalization and rounding up, E is selected as the next transfer destination.

FIG. 16 is a diagram showing an example of the connection line failure in FIG. 1, and FIG. 17 is a diagram showing an example of re-association of the network of FIG. 16 with the grid coordinate space. If the failure of the packet transfer relay device or connection line lasts for a long time, or if a failure has occurred or recovery from the failure has occurred at multiple points, the topology map device determines the physical connection topology that is effective at that time and the previous one. It is detected by the dynamic route control protocol or a notification from the device in the network that the association with the multi-dimensional coordinate space deviated from that of the multi-dimensional coordinate space, and the re-association with the multi-dimensional coordinate space is performed. FIG. 16 shows the re-association performed by the topology map device that has detected a long-term interruption of the connection line between the packet transfer relay devices G and Q shown in FIG. 16 in the topology of FIG.

[0032]

As described above, according to the present invention,
When the packet transfer relay device performs packet transfer,
Since the number of entries in the coordinate search table to be searched can be small regardless of the size of the network, there is a remarkable effect that high-speed search processing can be performed even in a large-scale network.

[Brief description of the drawings]

FIG. 1 is a diagram showing an example of a packet transfer network applied to the present invention.

FIG. 2 is a block diagram of a conventional packet transfer relay device.

FIG. 3 is a conceptual diagram of a routing table of a conventional packet transfer relay device.

FIG. 4 is a block diagram of an address map device showing one embodiment of the present invention.

FIG. 5 is a diagram showing an example of association of the network of FIG. 1 with a grid coordinate space.

FIG. 6 is a block diagram of a packet transfer relay device according to an embodiment of the present invention.

FIG. 7 is a diagram showing a concept of transfer by coordinates in the present invention.

FIG. 8 is an explanatory diagram of a transfer process using coordinates in the packet transfer relay device according to the embodiment (1) of the present invention.

FIG. 9 is an explanatory diagram of a transfer process based on coordinates in the packet transfer relay device according to the embodiment (2) of the present invention.

FIG. 10 is a diagram illustrating an example (1) of an information distribution range of local topology information according to the present invention.

FIG. 11 is a diagram illustrating an example (2) of an information distribution range of local topology information according to the present invention.

FIG. 12 is a diagram illustrating an example of a distribution range of local topology information in FIG. 5;

FIG. 13 is a diagram illustrating an example of a connection line failure in FIG. 5;

14 is a diagram illustrating an example (1) of a packet transfer relay process in the packet transfer relay device D in FIG. 13;

15 is a diagram illustrating an example (2) of the packet transfer relay process in the packet transfer relay device D in FIG. 13;

FIG. 16 is a diagram illustrating an example of a connection line failure in FIG. 1;

17 is a diagram showing an example of re-association of the network of FIG. 16 with a grid coordinate space.

[Explanation of symbols]

11: input port, 12: distribution unit, 13: routing control protocol processing unit, 14: normal transfer processing unit, 15: routing table,
16 output port, 21 inquiry processing unit, 22 topology mapping unit, 23 map storage unit, 17 local topology processing unit, 18 coordinate transfer table, 1
9: coordinate transfer processing unit, 20: coordinate label creation unit.

Claims (8)

[Claims] 1. A packet transfer relay device connected to a terminal and another packet transfer relay device via a network and performing packet transfer between them, wherein each packet transfer is performed based on a mutual connection status of the packet transfer relay devices. Make the relay device correspond to a point on the multidimensional coordinate space,
Means for acquiring coordinates of each point; means for calculating a distance in a multidimensional coordinate space between another packet transfer relay device serving as a transfer destination candidate and the packet transfer relay device of the transmission source; Means for selecting, as a transfer destination, a packet transfer relay device having the shortest distance. 2. The packet transfer relay device according to claim 1, further comprising: means for storing coordinates of each of said packet transfer relay devices in a multidimensional coordinate space. 3. The packet transfer relay device according to claim 1, further comprising: a unit for adding, to the packet to be transferred, information on coordinates in a multidimensional coordinate space of a packet transfer relay device of a transmission destination. A packet transfer relay device characterized by the above-mentioned. 4. A case in which each packet transfer relay device is associated with a point on a multidimensional coordinate space based on the mutual connection status of the packet transfer relay devices in the network, and a packet is transmitted from an arbitrary packet transfer relay device. When information about coordinates in the multi-dimensional coordinate space corresponding to the packet transfer relay device of the transmission destination is added to the packet, and when the packet is transmitted to the adjacent packet transfer relay device, one or more adjacent packets are relayed. Based on the coordinates of the packet transfer relay device and the coordinates of the packet transfer relay device of the transmission source to one of the packet transfer relay devices having the minimum distance in the multidimensional coordinate space of the packet transfer relay device of the transmission destination. A packet transfer control method, comprising: 5. The packet transfer control method according to claim 4, wherein each packet relay device in the network is associated with a point in a multidimensional coordinate space using a dynamic routing protocol or a topology. Each packet transfer relay device is associated with a point in the multidimensional coordinate space based on the connection information between each packet transfer relay device obtained from the setting information by the mapping unit, and the correspondence is stored in a map storage unit. In response to an inquiry including the address of a packet transfer relay device in a network, the address is searched by the map storage unit and the coordinates of the node obtained by the correspondence are answered by an address map device. Packet transfer control method. 6. The packet transfer control method according to claim 4, wherein each of the packet transfer relay devices in the network makes an inquiry to the address map device and relates to the packet transfer relay device in the network. When a packet is transferred, the source device of the packet or the packet transfer relay device directly accommodating the source device corresponds to the destination address with reference to the above-mentioned correspondence table. If the information on the coordinates of the transfer destination device to be transferred or the packet transfer relay device accommodating the transfer destination device exists in the above correspondence table, it is obtained by inquiring the address map device if it does not exist. Packet information, the label information relating to the coordinates of the destination device being added to the packet. Transfer control method. 7. The packet transfer control method according to claim 4, wherein the packet transfer relay device monitors a failure state of an adjacent packet transfer relay device or link.
When a failure occurs or recovers, a failure notification is performed to the packet transfer relay device whose coordinates are close to each other in the multidimensional coordinate space, and the notified packet transfer relay device determines a next transfer destination based on the label information. A packet transfer control method that uses the fault information to bypass a fault location. 8. The packet transfer control method according to claim 4, wherein when the degree of failure, such as the number of failures in the network or a suspension period, exceeds a threshold value, a dynamic state is determined. The address map device recognizes the situation by using a routing protocol or by notification from a packet transfer relay device near the failure, and the address map device connects each packet transfer relay device at that time. A packet transfer control method characterized by re-associating each packet transfer repeater with a point in a multidimensional coordinate space based on physical connection information of a connection line to be connected.