WO2012048615A1

WO2012048615A1 - Method and system for selecting relay nodes, and relay control node

Info

Publication number: WO2012048615A1
Application number: PCT/CN2011/080053
Authority: WO
Inventors: 李漓春
Original assignee: 中兴通讯股份有限公司
Priority date: 2010-10-14
Filing date: 2011-09-22
Publication date: 2012-04-19
Also published as: CN102447614A; CN102447614B

Abstract

Disclosed is a method for selecting relay nodes. A relay control node obtains from at least one network topology information server network mappings and cost mappings, and generates or updates a list of relay nodes according to the obtained relay nodes information; upon receiving from a first correspondent node a request for relay nodes, the relay control node processes the list of relay nodes according to the request for relay nodes, the network mappings, and the cost mappings, and returns the processed list of relay nodes to the first correspondent node; the first correspondent node selects a relay node for data forwarding from the received list of relay nodes. Also disclosed are a system for selecting relay nodes and a relay control node. By applying the method as put forth in the present invention, the selection of relay nodes can be reasonably achieved in a simple manner by using few messages, thereby alleviating the burden of network topology information servers.

Description

Relay node selection method, system and relay control node

The present invention relates to a relay technology in a network, and in particular, to a method, system, and relay control node for selecting a relay node. Background technique

In the Internet, communication between two peer nodes often encounters connectivity and quality of service (QoS) issues of IP layer transmission paths, including: Network Address Translation (NAT) devices or firewall devices can block peer nodes. The two parties establish a direct connection path for communication; because the Internet routing policy cannot consider the quality of service requirements of all communication parties, the quality of the direct connection path between the two peer nodes cannot meet the communication requirements; due to the huge Internet, the IP routing pair links The interrupt response is sluggish, causing a short-term interruption of the direct connection path between the two peer nodes, affecting communication; due to the defect of the routing protocol, some IP routing protocols cannot reflect some link congestion, resulting in the relationship between the two peer nodes. The direct connection path will have long-term congestion; wait.

For the above problems, a quick and effective solution is to use relay technology. The relay technology refers to a technology in which a relay node performs packet forwarding by adding one or more relay nodes to the IP packet delivery path. The relay can implement the NAT/firewall traversal function to help the peer nodes establish a direct connection; the relay can compensate for the lack of IP layer routing; the relay can bypass the broken and congested links.

According to the topology information of the network, a reasonable relay node is selected, and a relay path with high quality of service can be obtained, thereby improving the communication quality of both the peer nodes. The network topology information on which the relay node is selected may be from the measurement, the AS topology relationship extracted from the Border Gateway Protocol (BGP) routing table, and the network topology information server. The measurement overhead is large, and the topology information included in the BGP routing table is inaccurate and incomplete, and the network topology information server is used by the network. The network topology information on the network topology information server is maintained by the network operator, and the network topology information server can provide rich and accurate topology information for the application to optimize performance. Distribution of the Application Layer Traffic Optimization (ALTO) Working Group of the International Organization for Standardization Internet Engineering Task Force (IETF), ITU-T for ITU Telecommunication Standardization Sector The DSN (Distributed Service Network) working group is working on the standardization of the network topology information server access protocol.

At present, the existing scheme for using the network topology information server for relay node selection is the ALTO server proposed by the International Organization for Standardization (IETF) for the relay selection scheme. The network structure of the relay node selection using the network topology information server is as shown in FIG. 1. The peer node 1 is connected to the network topology information server 1, and the peer node 2 is connected to the network topology information server 2, After the node forwards the data packet, the method for using the network topology information server to perform relay node selection is as shown in FIG. 2, and the method includes the following steps:

Step 101: The peer node 1 requests the network topology information server 1 to request the relay node list, and sends the use relay notification to the opposite node 2;

Step 102: After receiving the relay notification, the peer node 2 requests the network topology information server 2 to request the relay node list.

Step 103: The network topology information server 2 sends its own relay node information to the network topology information server 1;

Step 104: The network topology information server 1 crosses the relay node of the network topology information server 2 with its own relay node according to the relay node information sent by the network topology information server 2, and obtains a relay node list, and Returning to the peer node 1 and the network topology information server 2; Step 105: The network topology information server 2 returns the relay node list to the peer node Step 106: The peer node 1 and the peer node 2 respectively measure the quality of service of the relay node according to the relay node list.

Step 107: The peer node 2 sends the measurement result of the quality of the relay node to the peer node 1;

Step 108: The opposite end node 1 selects an optimal relay node according to the measurement result of the service quality of the relay node of itself and the opposite end node 2.

In the prior art, another method for selecting a relay node by using a network topology information server, as shown in FIG. 3, the method includes the following steps:

Step 201: The peer node 1 sends a relay node list sorting request to the network topology information server 1, and sends a use relay notification to the peer node 2;

Step 202: After receiving the relay node list sorting request, the network topology information server performs sorting of the relay node list, and returns the sorted relay node list to the peer node 1;

Step 203: After receiving the relay notification, the peer node 2 sends a relay node list sorting request to the network topology information server 2;

Step 204: After receiving the relay node list sorting request, the network topology information server performs the sorting of the relay node list, and returns the sorted relay node list to the peer node 2;

Step 205: The peer node 1 and the peer node 2 respectively measure the service quality of the relay node according to the sorted relay node list.

Step 206: The peer node 2 sends the measurement result of the quality of the relay node to the peer node 1;

Step 207: The opposite end node 1 selects an optimal relay node according to the measurement result of the service quality of the relay node of the own node and the opposite end node 2.

In the above two existing solutions, each time the relay node is selected, the peer node 1 and the peer node 2 need to interact with the network topology information server 1 and the network topology information server 2, respectively. More, and easy to cause network topology information server 1, network topology information service The burden on the server 2 is too heavy, and the method is complicated and unreasonable. Summary of the invention

In view of this, the main object of the present invention is to provide a method, a system, and a relay control node for selecting a relay node, and reasonably implementing the selection of the relay node, the number of messages is small, and the method is simple.

In order to achieve the above object, the technical solution of the present invention is achieved as follows:

The invention provides a method for selecting a relay node, and the method includes:

The relay control node acquires the network mapping and the cost mapping from the at least one network topology information server, and the relay control node generates or updates the relay node list according to the obtained relay node information; the relay control node receives the first pair each time After the relay node of the end node requests, the relay node list is processed according to the relay node request, the network map, and the cost map, and the processed relay node list is returned to the first a pair of end nodes;

The first peer node selects relay node relay data from the received relay node list. In the above solution, the relay node request includes first peer node information and second peer node information;

The pair of end node information and the second pair of node information include: a node IP address of the pair of end nodes and the second peer node, and/or a node identifier, and/or a user identifier;

The first peer node is a peer node that initiates a relay transmission, and the second peer node is a destination peer node that sends data to the first peer node.

In the above solution, the relay node information includes: a node address of each relay node, and may further include a load condition, and/or a quality of service measurement result.

In the foregoing solution, the processing, by the relay node request, the network mapping, and the cost mapping, processing the relay node list, according to: a node IP according to the first peer node and the second peer node The IP prefix of the address is searched for a mapping of a Provider-defined Network Location Identifier ( PID) in the network mapping. When found, the relay control node obtains a PID set of all the relay nodes according to the network mapping: Relay PID1~Relay PIDn, and calculates a relay path cost relayed from each relay PID through a cost map, and processes the relay node list according to the relay path cost.

In the above solution, the processing is scoring and/or sorting and/or screening; wherein, when scoring at the cost of the relay path, the lower the cost, the higher the score; when sorting at the cost of the relay path, The lower the cost, the higher the ranking; when the cost of the relay path is used as the filtering condition, the relay node whose cost exceeds the threshold is screened out from the relay node list.

In the above solution, the selecting the relay node is: selecting the relay node with the highest score of the score, and/or selecting any one of the relay node lists after filtering, and/or selecting the sorting rank in the most The previous relay node.

The present invention provides a relay node selection system, the system includes: a relay control node, at least one network topology information server, a relay node, and a first pair of nodes;

a relay control node, configured to acquire a network mapping and a cost mapping from at least one network topology information server, and generate or update a relay node list according to the obtained relay node information; in each receiving the first peer node After the node request, the relay node list is processed according to the relay node request, the network mapping, and the cost mapping, and the processed relay node list is returned to the first peer node;

a network topology information server, configured to provide a network mapping and a cost mapping to the relay control node; and a relay node, configured to transit the data sent by the first peer node;

笫 A pair of end nodes, configured to send a relay node request to the relay control node, and select relay node relay data from the received relay node list.

In the above solution, the relay control node includes: a network topology information client unit, a relay processing unit, and a relay node information management unit;

a network topology information client unit, configured to obtain a network mapping and a cost mapping from at least one network topology information server;

a relay node information management unit, configured to generate or update a relay node according to the relay node information List

a relay processing unit, configured to process the relay node list according to the relay node request, the network mapping, and the cost mapping after each receiving a relay node request of the first peer node And returning the processed relay node list to the first peer node.

In the above solution, the relay processing unit is specifically configured to search for a mapping of the corresponding PID in the network mapping according to the IP prefix of the node IP address of the first peer node and the second peer node, and when found, according to The network map obtains a PID set of all relay nodes: a relay PID1~a relay PIDn, and calculates a relay path cost relayed from each relay PID through a cost map, and a relay node list according to the relay path cost Process it.

In the above solution, the relay processing unit is specifically configured to score and/or sort and/or filter the relay node list; wherein, when the cost of the relay path is scored, the cost is smaller. Higher; when sorting at the cost of the relay path, the lower the cost, the higher the ranking; when the cost of the relay path is used as the filtering condition, the relay with the cost exceeding the threshold is filtered out from the list of relay nodes. node.

In the foregoing solution, the first peer node is specifically configured to select a relay node with the highest score, and/or select any one of the filtered relay node lists, and/or select the sorting. The top relay node.

The present invention provides a relay control node, where the relay control node includes: a network topology information client unit, a relay processing unit, and a relay node information management unit;

a relay node information management unit, configured to generate or update a relay node list according to the obtained relay node information;

a relay processing unit, configured to: after each receiving a relay node request of the first peer node, according to the relay node request, the network mapping, and the cost mapping to the relay node column The table is processed, and the processed relay node list is returned to the first peer node.

The relay control node acquires the network mapping and the cost mapping from the at least one network topology information server, and the relay control node generates or updates the relay node list according to the obtained relay node information; the relay control node receives the relay node each time After the request, the relay node list is processed according to the relay node request, the network mapping, and the cost mapping, and the processed relay node list is returned to the proxy node of the first peer node;

The proxy node of the first peer node selects a relay node from the relay node list returned by the relay control node, and returns to the first peer node;

The first peer node relays data through the relay node.

In the foregoing solution, the processing, by the relay node request, the network mapping, and the cost mapping, processing the relay node list, according to: a node IP according to the first peer node and the second peer node The IP prefix of the address is searched for a mapping of the corresponding PID in the network mapping. When found, the relay control node obtains the PID set of all the relay nodes according to the network mapping: the relay PID1~the relay PIDn, and the cost mapping A relay path cost relayed from each relay PID is calculated, and the relay node list is processed according to the relay path cost.

The present invention provides a relay node selection system, the system comprising: a relay control node, at least one network topology information server, a relay node, a first pair of nodes, and a proxy node of the first pair of nodes; ,

a relay control node, configured to acquire a network mapping and a cost mapping from the at least one network topology information server 82, and generate or update a relay node list according to the obtained relay node information; after each receiving the relay node request And processing, according to the relay node request, the network mapping, and the cost mapping, the relay node list, and returning the processed relay node list to the proxy node;

a network topology information server, configured to provide a network mapping and a cost mapping to the relay control node; a relay node, configured to transit data sent by the first peer node;

a pair of end nodes, configured to request to the proxy node to send data to the second peer node; and transfer data through the relay node;

a proxy node, configured to send a relay node request to the relay control node according to the request of the first peer node; and select a relay node from the relay node list returned by the relay control node, and select the relay node information Return to the first peer node.

The invention provides a method, a system and a relay control node for selecting a relay node. The relay control node obtains a network map and a cost map from at least one network topology information server, and obtains a network map according to the obtained Relay node information generates or updates a relay node list; after each receiving a relay node request of a pair of end nodes, according to the relay node request, the Network map, and the Cost ma pair The relay node list is processed, and the processed relay node list is returned to the first peer node; the first peer node selects the relay node transit data from the received relay node list; The selection of the relay node is reasonable, the number of messages is small, and the method is simple. Moreover, each time the relay node is selected, the peer node does not need to interact with the network topology information server, and the network topology information server is mitigated. burden. DRAWINGS

1 is a schematic diagram of a network structure for selecting a relay node by using a network topology information server in the prior art;

2 is a schematic flowchart of a method for selecting a relay node by using a network topology information server in the prior art;

FIG. 3 is a schematic flowchart of another method for performing relay node selection by using a network topology information server in the prior art;

4 is a schematic flowchart of a method for implementing a relay node according to the present invention;

FIG. 5 is a schematic flowchart diagram of a specific implementation method of step 304 of the present invention; 6 is a schematic structural diagram of a system for implementing a relay node according to the present invention; FIG. 7 is a schematic flowchart of another method for selecting a relay node according to the present invention;

FIG. 8 is a schematic structural diagram of another selection system for implementing a relay node according to the present invention. detailed description

The basic idea of the present invention is: the relay control node acquires a Network map and a Cost map from at least one network topology information server, and generates or updates a relay node list according to the obtained relay node information; After the relay node of the end node requests, the relay node list is processed according to the relay node request, the network ma and the cost map, and the processed relay node list is returned to the first pair. End node; the first peer node selects relay node relay data from the received relay node list.

The invention will be further described in detail below with reference to the drawings and specific embodiments.

The present invention implements a method for selecting a relay node. As shown in FIG. 4, the method includes the following steps:

Step 301: The relay control node acquires an etwork map and a cost map from at least one network topology information server, and generates or updates a relay node list according to the relay node information.

Specifically, the relay control node requests a network map and a cost map to the at least one network topology information server by using a HyperText Transfer Protocol (HTTP) message, and after the at least one network topology information server receives the request, Returning a network map and a cost map to the relay control node, and the relay control node collects relay node information in advance, and generates or updates a relay node list;

The relay control node may obtain the latest Network map and the Cost map one or more times from a network topology information server, where the acquired Network map and Cost map may be used for multiple relay node selection; The relay node information collected by the control node may be configured by the administrator to configure the relay node information to the relay control node, or may actively report the relay node information to the relay control node when the relay node goes online. Obtained, the relay The node information includes a node address of each relay node, and may also include a load condition, a quality of service measurement result, and the like, and may be used for multiple relay node selection;

The Network ma includes a mapping of an IP prefix of all nodes to a network ID, that is, a PID; the Cost ma includes a path cost value between the PIDs, and all the nodes include a peer node and a relay node.

Step 302: The first peer node sends a relay node request to the relay control node each time the first peer node needs to send data to the second peer node.

Here, the relay node request includes first peer node information and second peer node information; the first peer node information and the second peer node information include a first peer node and a second peer node Node IP address, and/or node identification, and/or user identification, etc.;

Step 303: The relay control node receives and accepts a relay node request sent by the first peer node each time;

Step 304: The relay control node processes the relay node list according to the first and second peer node IP addresses, the network map, and the cost ma.

Specifically, the relay control node searches for a mapping association of the corresponding PID in the network ma according to the IP prefix of the node IP address of the first peer node and the second peer node, and if found, the relay control node according to the network map Obtaining PID sets of all relay nodes: relay PID1~relay PIDn, and calculating a relay path cost relayed from each relay PID through Cost ma, and processing the relay node list according to the relay path cost If not found, the relay control node re-acquires the Network map and the Cost map from at least one network topology information server, and performs this step again;

In this step, if the relay node request does not include the node IP address of the first peer node and the second peer node, the relay control node needs to send the first node and the second node in the relay node request. Converting the node identifier or user identifier of the peer node to the node IP address of the first peer node and the second peer node;

The processing may be scoring and/or sorting and/or screening; wherein, when scoring at the cost of the relay path, the lower the cost, the higher the score, and the higher the score, the better the relay node; When sorting at the cost of the relay path, the lower the cost, the higher the ranking, that is, the better the relay node; when the cost of the relay path is used as the screening condition, the cost of screening from the list of relay nodes exceeds the pre- The relay node of the set threshold.

Further, the scoring, sorting, and screening may also consider other factors such as the load status of the relay node, and/or the quality measurement result of the J service.

Step 305: The relay control node returns the processed relay node list to the first peer node. Step 306: After obtaining the processed relay node list, the first peer node selects the relay node to transfer the data to the second pair. End node:

In this step, the selecting the relay node is generally selecting the relay node with the highest score of the score, and/or selecting any one of the relay nodes in the filtered relay node list, and/or selecting the sorting rank in the most The relay node in front; when there are multiple relay nodes with the highest score, the relay node can be arbitrarily selected;

The data is first sent by the first peer node to the relay node, and then sent by the relay node to the second peer node; similarly, the second peer node may also use the relay node to transfer data to the first peer end. node.

In the foregoing step 301, the relay control node acquires the network map, and the order of the cost map and the relay control node to collect the relay node information may be arbitrary, and may be performed simultaneously.

For the specific implementation method of step 304, as shown in FIG. 5, the following steps are included: Step 304a: The relay control node performs preliminary screening on the relay node according to other factors such as the load of the relay node;

Here, the relay node is generally listed with a load greater than a preset threshold from the relay node list. In the delete, this step is optional.

Step 304b: The relay control node obtains the PID set of all the relay nodes according to the etwork map: relay PID1 ~ relay PIDn;

In this step, the relay control node may maintain and update the PID set of all the relay nodes as the relay node joins and exits, wherein one relay PID will correspond to at least one relay node.

Step 304c: The relay control node searches for the PID corresponding to the IP addresses of the first and second peer nodes according to the Network map: the first peer PID and the second peer PID;

Step 304d: The relay control node calculates a relay path cost for relaying from each relay PID. In this step, the relay path cost may adopt a forward relay path cost, a reverse relay path cost, or a two-way medium. Following the path cost, the calculation of the three requires the use of a Cost map, where cost ( PID A, PID B ) refers to the value of the direct path from PID A to PID B. Forward relay path cost = cost (first peer PID, relay PIDi) + cost (relay PIDi, second peer PID); reverse relay path cost =. (^ (second peer PID, relay PIDi) + cost (relay PIDi, first peer PID); two-way relay path cost = forward relay path cost + reverse relay path cost.

Step 304e: The relay control node scores and/or filters and/or sorts the relay nodes corresponding to each relay PID according to the relay path cost.

In this step, when scoring with the relay path cost, the lower the cost, the higher the score; when sorting with the relay path cost, the more the cost, the higher the ranking; the relay path When the cost is used as a screening condition, the relay node whose cost exceeds a preset threshold is screened out from the relay node list.

Wherein, when the relay node corresponding to each relay PID is scored, the scores of the relay nodes corresponding to the same relay PID are the same; when the relay nodes corresponding to each relay PID are sorted, the corresponding The phase of the relay node of a relay PID.

Further, the step further includes: the relay control node combines the relay path cost into each of the The relay nodes corresponding to each relay PID are scored and/or filtered and/or ordered according to other factors such as the load condition of the node, and/or the quality of service measurement.

In order to implement the above method, the present invention further provides a selection system for implementing a relay node. As shown in FIG. 6, the system includes: a relay control node 61, at least one network topology information server 62, a relay node 63, and a first Peer node 64; where

a relay control node 61, configured to acquire a network map and a cost map from at least one network topology information server 62, and generate or update a relay node list according to the obtained relay node information; each time receiving the first peer node After the relay node requests 64, the relay node list is processed according to the relay node request, the network map, and the cost map, and the processed relay node list is returned to the pair of terminals. Node 64;

a network topology information server 62, configured to provide an etwork map and a cost map to the relay control node 61;

a relay node 63, configured to transfer data;

The first peer node 64 is configured to send a relay node request to the relay control node 61, and select the relay node 63 to transfer data from the received relay node list;

The relay control node 61 includes: a network topology information client unit 611, a relay node information management unit 612, and a relay processing unit 613;

a network topology information client unit 611, configured to obtain an ear and a network ma and a cost map from at least one network topology information server 62;

a relay node information management unit 612, configured to generate or update a relay node list according to the obtained relay node information;

The relay processing unit 613 is configured to process the relay node list according to the relay node request, the Network ma, and the Costa map after each receiving the relay node request of the first correspondent node 64, and Returning the processed relay node list to the first peer node 64;

The relay node request includes first peer node information and second peer node information; The first peer node information and the second peer node information include: a node IP address of the first peer node 64 and the second peer node, and/or a node identifier, and/or a user identifier, and the like;

The relay node information includes a node address of each relay node 63, and may further include: a load condition, and/or a quality of service measurement result, etc., which may be used for multiple relay node selection; Generally, the relay node information may be configured by the administrator to the relay control node, or may be obtained by actively reporting the relay node information to the relay control node when the relay node goes online.

The relay processing unit 613 processes the relay node list according to the relay node request, the network map, and the core map, specifically: the relay processing unit 613 is configured according to the first peer node 64 and the second The IP prefix of the node IP address of the peer node searches for a mapping of the corresponding PID in the Network map. If found, the relay processing unit 613 obtains the PID set of all the relay nodes 63 according to the Network map: relay PID1~relay PIDn, and calculating a relay path cost for relaying from each relay PID through a Costa map, processing the relay node list according to the relay path cost; if not, notifying the network topology information client unit 611 re-acquiring the network map and the cost map from the at least one network topology information server 62, and the relay processing unit 613 searches for the mapping in the network map again;

Further, the relay processing unit 613 is further configured to combine the relay path cost with the load condition of each relay node 63, and/or other factors such as the quality of service measurement result, and corresponding to each relay PID. Following the node 63 scores and/or filters and/or sorts.

Further, the relay path cost is a forward relay path cost, and is: a first pair of end sections Point 64 to the relay node 63 path cost and the sum of the relay node 63 to the second peer node path cost; or, the relay path cost is the reverse relay path cost, which is: the second peer node to the middle Following the path cost of the node 63 and the path cost of the relay node 63 to the first correspondent node 64; or, the relay path cost is a two-way relay path cost, which is: a forward relay path cost and a reverse relay The sum of the path costs.

The first peer node 64 selects a relay node from the received relay node list, specifically: the first peer node 64 selects the relay node with the highest score, and/or selects the filtered Any one of the relay nodes in the relay node list, and/or select the top ranked relay node; when there are multiple relay nodes with the highest score, you can arbitrarily select one of the relays node.

Based on the foregoing system, the present invention further provides a relay control node. As shown in FIG. 6, the relay control node 61 includes: a network topology information client unit 611, a relay node information management unit 612, and a relay processing unit. 613; Among them,

a network topology information client unit 611, configured to obtain a network map and a cost map from at least one network topology information server 62;

The relay processing unit 613 is configured to: after each receiving the relay node request of the first peer node 64, the relay node list according to the relay node request, the network map, and the cost map Processing, and returning the processed relay node list to the first peer node 64;

The relay node request includes first peer node information and second peer node information; the first peer node information and the second peer node information include: a first peer node 64 and a second peer node Node IP address, and/or node identification, and/or user identification, etc.;

The relay node information includes a node address of each relay node 63, and may further include: The load condition, and/or the quality of service measurement result, etc., may be used for multiple relay node selection; the relay processing unit 613 performs the relay node list according to the relay node request, the Network map, and the Cost map. The processing is specifically: the relay processing unit 613 searches for a mapping of the corresponding PID in the Network map according to the IP prefix of the node IP address of the first peer node 64 and the second peer node, and if found, relays The processing unit 613 obtains the PID set of all the relay nodes 63 according to the Network map: the relay PID1~the relay PIDn, and calculates the relay path cost relayed from each relay PID through the Costa map, according to the relay path cost Processing the relay node list; if not found, notifying the network topology information client unit 611 to reacquire the network map and the cost map from the at least one network topology information server 62, the relay processing unit 613 re-establishing in the Network Find the mapping in the map;

Further, the relay processing unit 613 is further configured to combine the relay path cost with the load condition of each relay node 63, and/or other factors such as the quality of service measurement result, and correspond to each relay PID. Following the node 63 scores and/or filters and/or sorts.

Further, the relay path cost is a forward relay path cost, which is: a sum of a path cost of the pair of end nodes 64 to the relay node 63 and a path cost of the relay node 63 to the second opposite end node; or The relay path cost is a reverse relay path cost, which is: a sum of a path cost of the second correspondent node to the relay node 63 and a path cost of the relay node 63 to the first correspondent node 64; or, the relay The path cost is the two-way relay path cost, which is: the sum of the forward relay path cost and the reverse relay path cost.

The present invention also provides a method for selecting a relay node, as shown in FIG. 7, the method package Including the following steps:

Step 701: The relay control node acquires an etwork map and a cost map from at least one network topology information server, and generates or updates a relay node list according to the relay node information.

Step 702: The first peer node sends a relay node request to the relay control node by using the proxy node each time the first peer node needs to send data to the second peer node.

Specifically, the first peer node requests its proxy node to send data with the second peer node; the proxy node sends a relay node request to the relay control node according to the request of the first peer node;

The request of the first peer node and the request of the relay node both include the first peer node information and the second peer node information; the first peer node information and the second peer node information include the first peer end The node IP address of the node and the second peer node, and/or the node identifier, and/or the user identity, and the like.

Step 703: The relay control node receives and accepts the relay node request of the first peer node. Step 704: The relay control node selects the relay node according to the first and second peer node IP addresses, the Network map, and the Cost ma. List processing;

Specifically, the relay control node searches for a mapping association of the corresponding PID in the network ma according to the IP prefix of the node IP address of the first peer node and the second peer node, and if found, the relay control node according to the network map Obtaining PID sets of all relay nodes: relay PID1~relay PIDn, and calculating a relay path cost relayed from each relay PID through Cost ma, and processing the relay node list according to the relay path cost If not found, the relay control node re-executes the etwork map and the cost map from at least one network topology information server, and performs this step again;

In this step, if the relay node request does not include the node IP address of the first peer node and the second peer node, the relay control node needs to send the first node and the second peer in the relay node request. The node identifier or user ID of the node is converted into a section of the first peer node and the second peer node Point IP address;

Further, the scoring, sorting, and screening may also consider other factors such as the load status of the relay node, and/or the quality of service measurement.

Step 705: The relay control node returns a processed relay node list to the proxy node of the first peer node.

Step 706: The proxy node of the first peer node selects a relay node from the relay node list returned by the relay control node.

Step 707: The proxy node of the pair of end nodes returns the selected relay node to the first peer node.

Step 708: The first peer node transfers the data to the second peer node by using the relay node. In this step, the first peer node obtains one or more relay nodes, and when multiple relay nodes are obtained, One of the relay nodes may be further selected to transfer data to the second peer node.

In the foregoing step 707, the proxy node of the first peer node may also not select the relay node, and forward the relay node list returned by the relay control node to the first peer node;

Correspondingly, in the foregoing step 708, after obtaining the processed relay node list, the first peer node selects the relay node to transfer the data to the second peer node.

The present invention also provides a selection system for implementing a relay node. As shown in FIG. 8, the system includes: a relay control node 81, at least one network topology information server 82, a relay node 83, and a first peer node 84. a proxy node 85 of the first pair of end nodes 84; wherein

a relay control node 81, configured to obtain from at least one network topology information server 82 a network map and a cost map, and generating or updating a relay node list according to the obtained relay node information; and each time receiving the relay node request, according to the relay node request, the network map, and the Cost map Processing the relay node list, and returning the processed relay node list to the proxy node 85;

a network topology information server 82, configured to provide a network map and a cost map to the relay control node 81;

a relay node 83, configured to transfer data;

The first peer node 84 is configured to request its proxy node 85 to send data with the second peer node; and transfer data through the relay node 83;

The proxy node 85 is configured to send a relay node request to the relay control node according to the request of the first peer node 84; and select the relay node 83 from the relay node list returned by the relay control node, and select the relay The information of the node 83 is returned to the first peer node 84;

The relay control node 81 includes: a network topology information client unit 811, a relay node information management unit 812, and a relay processing unit 813;

a network topology information client unit 811, configured to obtain a network map and a cost map from at least one network topology information server 62;

a relay node information management unit 812, configured to generate or update a relay node list according to the obtained relay node information;

The relay processing unit 813 is configured to process the relay node list according to the relay node request, the network map, and the core map after receiving the relay node request, and process the processed relay node. The list is returned to the proxy node 85;

The relay node request includes first peer node information and second peer node information; the first peer node information and the second peer node information include: a first peer node 84 and a second peer node Node IP address, and/or node identification, and/or user identification, etc.;

The relay node information includes a node address of each relay node 83, and may further include: The load condition, and/or the quality of service measurement result, etc., may be used for multiple relay node selection; the obtaining relay node information, generally, the administrator may configure the relay node information into the relay control node, It may also be obtained by actively reporting the relay node information to the relay control node when the relay node goes online.

The relay processing unit 813 processes the relay node list according to the relay node request, the network ma, and the core map, specifically: the relay processing unit 813 is configured according to the first peer node 84 and the second The IP prefix of the node IP address of the peer node searches for a mapping of the corresponding PID in the network map. If found, the relay processing unit 813 obtains the PID set of all the relay nodes 83 according to the Network map: relay PID1~relay PIDn, and calculating a relay path cost for relaying from each relay PID through a Costa map, processing the relay node list according to the relay path cost; if not, notifying the network topology information client unit 811 re-acquiring network ma and Cost map from at least one network topology information server 82, and the relay processing unit 813 searches for the mapping in the network map again;

Further, the relay processing unit 813 is further configured to combine the relay path cost with the load condition of each relay node 83, and/or other factors such as the quality of service measurement result, and corresponding to each relay PID. Following node 83 scores and/or filters and/or sorts.

Further, the relay path cost is a forward relay path cost, which is: a sum of a path cost of the first peer node 84 to the relay node 83 and a path cost of the relay node 83 to the second peer node; or The relay path cost is a reverse relay path cost, which is: a path cost of the second peer node to the relay node 83 and a path cost of the relay node 83 to the first peer node 84; Alternatively, the relay path cost is a two-way relay path cost, which is: a sum of a forward relay path cost and a reverse relay path cost.

The method of the present invention can reasonably implement the selection of the relay node, the number of messages is small, and the method is simple. Moreover, each time the relay node is selected, the peer node does not need to interact with the network topology information server. Reduce the burden on the network topology information server.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included. Within the scope of protection of the present invention.

Claims

Claim

A method for selecting a relay node, the method comprising:

The relay control node acquires the network mapping and the cost mapping from the at least one network topology information server, and the relay control node generates or updates the relay node list according to the obtained relay node information; the relay control node receives the first pair each time After the relay node of the end node requests, the relay node list is processed according to the relay node request, the network mapping, and the cost mapping, and the processed relay node list is returned to the a pair of end nodes;

The first peer node selects relay node relay data from the received relay node list.

The selection method according to claim 1, wherein the relay node request includes first peer node information and second peer node information;

The first peer node information and the second peer node information include: a node IP address of the first peer node and the second peer node, and/or a node identifier, and/or a user identifier;

The selection method according to claim 1, wherein the relay node information comprises: a node address of each relay node, and may further include a load condition, and/or a quality of service measurement result.

The selection method according to claim 2, wherein the processing the relay node list according to the relay node request, the network mapping, and the cost mapping is: according to the first The IP prefix of the node IP address of the peer node and the second peer node searches for a mapping of the corresponding network location identifier (PID) in the network mapping. When found, the relay control node obtains all the relays according to the network mapping. The PID set of the node: relay PID1 ~ relay PIDn, and calculate the relay path cost relayed from each relay PID through the cost map, and process the relay node list according to the relay path cost.

The selection method according to claim 1 or 4, wherein the processing is Points and/or sorting and/or screening; wherein, when scoring at the cost of the relay path, the lower the cost, the higher the score; when sorting at the cost of the relay path, the lower the cost, the lower the ranking Before; when the cost of the relay path is used as a screening condition, the relay node whose cost exceeds the threshold is screened out from the relay node list.

The selection method according to claim 5, wherein the selecting the relay node is: selecting a relay node with the highest score of the score, and/or selecting any one of the filtered relay node lists. The relay node, and/or the selection of the top ranked relay node.

A relay node selection system, the system comprising: a relay control node, at least one network topology information server, a relay node, and a first pair of nodes; wherein, the relay control node uses Obtaining a network mapping and a cost mapping from at least one network topology information server, and generating or updating a relay node list according to the obtained relay node information; each time receiving the relay node request of the first peer node, according to The relay node request, the network mapping, and the cost mapping process the relay node list, and return the processed relay node list to the first peer node;

The first peer node is configured to send a relay node request to the relay control node, and select the relay node transit data from the received relay node list.

The selection system according to claim 7, wherein the relay control node comprises: a network topology information client unit, a relay processing unit, and a relay node information management unit;

a relay node information management unit, configured to generate or update a relay node list according to the relay node information; a relay processing unit, configured to process the relay node list according to the relay node request, the network mapping, and the cost mapping after each receiving a relay node request of the first peer node And returning the processed relay node list to the first peer node.

The selection system according to claim 8, wherein the relay processing unit is specifically configured to use an IP prefix according to a node IP address of the first peer node and the second peer node in the network mapping. Find the mapping of the corresponding PID. When finding, obtain the PID set of all relay nodes according to the network mapping: relay PID1~relay PIDn, and calculate the relay path cost relayed from each relay PID through the cost mapping, Processing the relay node list according to the relay path cost;

The selection system according to claim 9, wherein the relay processing unit is specifically configured to score and/or sort and/or filter the relay node list; When the cost of the path is scored, the lower the cost, the higher the score; when sorting at the cost of the relay path, the lower the cost, the higher the ranking; when the cost of the relay path is used as the filtering condition, the secondary relay The node list is filtered out of relay nodes whose cost exceeds the threshold.

The selection system according to claim 10, wherein the first peer node is specifically configured to select a relay node with the highest score, and/or select a filtered relay node list. Any of the relay nodes, and/or select the top ranked relay node.

A relay control node, the relay control node includes: a network topology information client unit, a relay processing unit, and a relay node information management unit;

a relay node information management unit, configured to generate or update a relay node list according to the obtained relay node information; a relay processing unit, configured to process the relay node list according to the relay node request, the network mapping, and the cost mapping after each receiving a relay node request of the first peer node And returning the processed relay node list to the first peer node.

13. A method for selecting a relay node, the method comprising:

The first peer node relays data through the relay node.

The method according to claim 13, wherein the processing of the relay node list according to the relay node request, the network mapping, and the cost mapping is: according to the first The IP prefix of the node IP address of the peer node and the second peer node searches for a mapping of the corresponding PID in the network mapping. When found, the relay control node obtains the PID set of all the relay nodes according to the network mapping: Relay PID1 ~ relay PIDn, and calculate a relay path cost relayed from each relay PID through a cost map, and process the relay node list according to the relay path cost.

A system for selecting a relay node, the system comprising: a relay control node, at least one network topology information server, a relay node, a first pair of nodes, and a proxy node of the first pair of nodes ; among them,

a relay control node, configured to acquire a network mapping and a cost mapping from the at least one network topology information server 82, and generate or update a relay node list according to the obtained relay node information; after each receiving the relay node request According to the relay node request, the network mapping, and the The cost mapping processes the relay node list, and returns the processed relay node list to the proxy node;

a first peer node, configured to request, to the proxy node, the data to be sent by the second peer node; and transfer the data through the relay node;