WO2012065506A1

WO2012065506A1 - Relay selection method and relay control system

Info

Publication number: WO2012065506A1
Application number: PCT/CN2011/081343
Authority: WO
Inventors: 王炜; 胡永生; 孟昱
Original assignee: 中兴通讯股份有限公司
Priority date: 2010-11-17
Filing date: 2011-10-26
Publication date: 2012-05-24
Also published as: CN102469017A

Abstract

Disclosed is a relay selection method. A relay control system collects a piece of relay service information, and forms a relay service history record. The relay control system processes the relay service history record, and acquires an effective relay service history record. When a user requests for a relay service, the relay control system selects a relay node on the basis of the effective relay service history record. Correspondingly disclosed is the relay control system. In the present invention, as a judgment can be made directly on the basis of previous history records during a relay selection, the need for a peer-to-peer service quality assessment is obviated, channel and bandwidth consumption for the relay selection is thereby reduced, and the speed of the relay selection is also accelerated.

Description

Relay selection method and relay control system

The present invention relates to network communication technologies, and in particular, to a relay selection method and a relay control system. Background technique

With the development of network technologies, multimedia services (including VoIP services, video on demand/live broadcast, software downloads, etc.) have rapidly spread in packet data networks. Especially in a network based on peer-to-peer technology, each node serves as a client to obtain services from other nodes, and also serves as a server to other nodes. When the peer node is behind a network address translation (NAT) device or a firewall device, the two parties cannot establish a connection directly. In this case, relay technology is needed to implement the NAT/firewall traversal function to help the peer establish a direct connection. Further, since the packet data network uses the best effort to transmit data packets, the quality of service is difficult to meet the requirements of multimedia services. Therefore, it is necessary to apply the relay technology in the packet data network. Strip transmission paths to improve network service quality.

The relay technology refers to a technique in which a packet is forwarded by a relay node by adding one or more relay nodes to a packet packet transmission path. Compared to traditional IP routing technologies, the advantages of overlay network-based relay technology are:

The relay technology can implement the NAT/firewall traversal function, which provides data transfer function when the two parties cannot directly establish a connection.

The relay technology can provide the IPv4/IPv6 protocol conversion function, that is, protocol conversion and data transfer function when the communication parties use different IP protocols.

- The relay technology can change the data transmission path. That is, when the default path cannot meet the quality of service requirements, the quality of service can be improved by adding a relay and changing the data transmission path. This The quality of service includes transmission delay, packet loss rate, transmission bandwidth, and delay jitter. The relay technology can provide multiple optional transmission paths, that is, when the default path is unreliable, when the transmission fails, the service continuity can be ensured by quickly switching the transmission path.

In view of these advantages of relay technology, relay technology is widely used in packet data networks. However, the relay technology needs to use a suitable relay node to improve the quality of service of the transmission. The wrong relay node often seriously affects the quality of service. The relay selection algorithm needs to select suitable nodes from a large number of candidate relay nodes. The current selection method is often based on relay groups. The basic idea is: to make relay nodes and common network nodes according to their IP network segments. Or the network topology is pre-assigned to a specific group. FIG. 1 is a schematic diagram of a group organization structure of the prior art. As shown in FIG. 1 , the group may include both a relay node and a common network enjoying relay services in the network. The nodes, such as group 110 in FIG. 1, include both relay nodes 111, 112 and normal network nodes 113, 114. The relay system measures the quality of service of the link between the relay groups, and the measurement content may include return delay (RTT), delay jitter, bandwidth, and the like.

FIG. 2 is a schematic flowchart of an existing group-based relay selection method. As shown in FIG. 2, the method includes:

Step 201: The first network node (source node) needs to use the relay service to communicate with the second network node (the destination node), and the first network node sends the address of the communication party to the relay control node to request the relay resource.

Step 202: After receiving the address information of the first and second network nodes, the relay control node obtains the group IDs of the first and second network nodes by using the network address and the group correspondence. Selecting one or more candidate groups according to the group ID of the communication parties and the quality of service measurement information between the groups (which can be measured in advance and stored in the relay control node) or network topology information, In the list of candidate relay groups.

Step 203: When the candidate relay group list includes multiple relays or multiple groups, the relay control node may request the group agent to perform actual measurement on the actual service quality information, and then narrow the candidate. The range of relay nodes to determine the final relay node.

Step 204: The relay control node interacts with the selected relay node A to reserve a relay resource.

Step 205: The relay control node returns the information of the relay node A to the first network node. Step 206: The first network node and the second network node establish communication through the relay node A. However, the following drawbacks exist in the prior art process:

1) In steps 202 and 203, the group measurement record granularity used for the relay selection reference is generally coarse, and can only serve as a reference for relay selection. Further measurement is often required during the relay selection process, which results in a longer selection process and a certain network bandwidth. Further, when the number of groups is large and the group measurement data is large, the relay selection process often needs to traverse all possible relay groups for selection, and the calculation amount is very large.

2) Group policies configured through the network lack flexibility in practical applications. Since the network status is subject to change, groups grouped by network address often cannot reflect the actual network status of the user. When the group partition is too large, users in the same group may have different network environments. Even if the same relay is used, the quality of service obtained may vary greatly. In this case, it is inevitable that relaying is required. After the group is selected, the quality of service of individual nodes and relays is measured separately in the group, and the measurement consumption is huge. Summary of the invention

In view of this, the main objective of the embodiments of the present invention is to provide a relay selection method and a relay control system, which can reduce the calculation amount of relay selection, save the relay selection time and system resources, and thereby improve the relay selection efficiency.

To achieve the above objective, the technical solution of the embodiment of the present invention is implemented as follows:

A relay selection method includes:

The relay control system collects relay service information to form a relay service history record;

The relay control system processes the relay service history record to obtain a valid relay service History

When the user applies for the relay service, the relay control system selects the relay node based on the valid relay service history record.

The relay service history record includes one or more of the following: a relay node involved in the relay service, a source node, an address/group ID of the destination node, a quality of service information of the relay service, and a time stamp of the relay service .

The relay service history record further includes one or more of the following: a relay node, a source node, and an access mode of the destination node involved in the relay service; and a service type of the relay service.

The quality of service is the quality of service between the source node and the destination node, or the quality of service between the source node/destination node and the relay node. The quality of service is manifested by one or more of the following: delay, Delay jitter, bandwidth, and packet loss rate.

The relay control system collects relay service information as:

The relay control system obtains the relay service information that is actively reported by the relay node, the source node, or the destination node, or

The relay control system requests relay service information from the relay node, the source node, or the destination node, or

The relay control system requests relay service information from other network information servers, or the relay control system acquires relay service information actively pushed by other network information servers. The relay control system saves the relay service information as: the relay control node in the relay control system is a single centralized server, and the relay service information is stored in the server, the related disk array or the network storage; After there are multiple peer-to-peer relay control nodes in the control system, the relay service information is distributed on different relay control nodes in a distributed storage manner.

The relay control system processes the relay service history record, and obtaining a valid relay service history record includes one or more of the following:

Delete expired history based on the timestamp of the history; Sort and aggregate historical records;

Keep records of the best quality of service in the same type of history.

When the user applies for the relay service, the relay control system selects the relay node according to the valid relay service history record as:

The source address/group ID and the destination address/group ID in the valid relay service history record are the same as the source address/group ID and the destination address/group ID respectively involved in the user requesting the relay service. The relay service history records the relay node described in the relay service history record as a candidate relay node.

The method further includes: the relay control system adjusts the group classification information according to the relay service history, specifically:

When the difference in quality of service obtained by the nodes in the group using the same relay node is not less than a preset value, the group is subdivided into a plurality of groups; when nodes in different groups use the same relay node When the difference in the obtained quality of service is not greater than a preset value, the different groups are merged.

A relay control system, comprising: a relay service information collection module, a relay service history record formation module, a relay service history record processing module, and a relay node selection module; wherein the relay service information collection module is configured For collecting relay service information;

The relay service history record forming module is configured to form a relay service history record according to the relay service information collected by the relay service information collection module;

The relay service history processing module is configured to process a relay service history record formed by the relay service history forming module to obtain a valid relay service history record; and the relay node selection module is configured to be When the user applies for the relay service, the relay node is selected according to the effective relay service history recorded by the relay service history processing module.

The relay service information collection module collects the relay service information as:

The relay control system acquires the relay service actively reported by the relay node, the source node, or the destination node. Information, or,

The relay control system requests relay service information from other network information servers, or the relay control system acquires relay service information actively pushed by other network information servers. The relay service history forming module forms the relay service information as: storing the relay service information on the server, the related disk array or the network storage, or storing the relay service information in a distributed storage manner.

The relay service history processing module processes the relay service history record, and obtaining a valid relay service history record includes one or more of the following:

Delete expired history based on the timestamp of the history;

Sort and aggregate historical records;

Keep records of the best quality of service in the same type of history.

When the user applies for the relay service, the relay node selection module selects the relay node according to the effective relay service history record processed by the relay service history processing module:

The system further includes a group classification information adjustment module, configured to adjust the group classification information according to the relay service history record, specifically:

When the difference in quality of service obtained by the nodes in the group using the same relay node is not d, at a preset value, the group is subdivided into multiple groups; when nodes in different groups use the same relay When the difference in quality of service obtained by the node is not greater than a preset value, the different groups are merged.

The relay selection method and the relay control system according to the embodiment of the present invention utilize the relay service calendar The history record assists the relay selection. According to the embodiment of the present invention, since the relay selection can directly make a judgment according to the historical history, the end-to-end quality of service measurement is not required, thereby reducing the signaling of the relay selection. And bandwidth consumption, but also speed up the relay selection speed. Further, the amount of calculation required according to the past history selection is smaller than traversing all possible relay groups, and the calculation amount of the relay selection algorithm can be reduced; in addition, the group of nodes (including the relay node and the source destination node) Groups can be dynamically optimized based on historical records so that group settings and actual network conditions can be more consistent. DRAWINGS

1 is a schematic diagram of a group organization structure of the prior art;

2 is a schematic flow chart of an existing group-based relay selection method;

3 is a schematic flowchart of a relay selection method according to an embodiment of the present invention;

4 is a schematic flowchart of a method for collecting service quality information according to an embodiment of the present invention; FIG. 5 is a schematic flowchart of another method for collecting service quality information according to an embodiment of the present invention; FIG. 6 is a schematic diagram of another method for collecting service quality according to an embodiment of the present invention; Schematic diagram of a method flow of information; FIG. 7 is a schematic flowchart of relay selection using a relay service history record according to an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of a relay selection system according to an embodiment of the present invention. detailed description

The basic idea of the invention is to use relay service history records to assist in relay selection.

In practical applications, the relay control node often performs a large number of relay selection tasks, and many of the communication source destination nodes have similar network conditions. Therefore, a successful relay selection can provide a reference for subsequent relay selection of similar requests. This can greatly reduce the time, bandwidth and computational consumption of the relay selection; on the other hand, even if the network address is different, the user may still be in the same physical network, and these users can belong to the same group, so that Reduce the number of groups and increase the efficiency of relay selection.

Based on the above considerations, the embodiment of the present invention provides a relay selection method based on a relay service history record. FIG. 3 is a schematic flowchart of a relay selection method according to an embodiment of the present invention. As shown in FIG. 3, the method includes :

Step 301: The relay control system collects relay service information to form a relay service history record. Generally, each relay service history record corresponds to one relay service, and the relay service history record may include one or more of the following: a relay node, a source node, and an address/group ID of the destination node involved in the relay service. The quality of service information of the relay service; the timestamp of the relay service, where the quality of service can be embodied by one or more of the following: delay, delay jitter, bandwidth, packet loss rate, etc. In addition, the quality of service can be It is the quality of service from end to end (between source node and destination node), and it can also be the quality of service between the end (source node/destination node) and the relay node. When the service time ratio is long, the service quality measurement value and the specific measurement time point of multiple time points can be retained in a single quality of service history.

It should be noted that the relay service history record may further include one or more of the following: a relay node, a source node, and an access mode of the destination node involved in the relay service; a service type of the relay service; Information following the service history (corresponding to the case where multiple relay nodes are used by this service).

It should be noted that after the relay control system obtains the relay service information such as the service quality information, the relay service information may be stored in a specific history database, and the specific storage manner may adopt multiple manners, when the relay control system When the relay control node is a single centralized server, the relay service information can be stored in the server, the related disk array or the network storage; when there are multiple peer relay control nodes in the relay control system, The service information can be distributed on different relay control nodes in a distributed storage manner. The specific implementation can implement storage and query by using a distributed hash table.

Step 302: The relay control system processes the relay service history record described in step 301. Filter out valid relay service history (for example, records with better quality of service) for subsequent relay selection.

Here, the manner in which the relay control system processes the relay service history may be implemented by a single centralized server or by multiple distributed servers. The specific processing may include at least one or more of the following, but is not limited to:

Delete expired history based on the timestamp of the history. Specifically, the system may set a specific expiration time limit to delete the expired history record;

The history records are classified and aggregated based on relevant information in the history. For example, the system may classify the history according to one or more combinations of the first network node, the IP address of the second network node, the network segment, the group ID, and the access mode, and will have the same source group and destination. The history of the group (or other information) is placed in a class; the system can also classify or further subdivide the history according to the type of the business, the time period of the business, etc.; for the same type of record, the system can perform data compression , merge relevant information to save storage space;

Keep records of the best quality of service in the same type of history. For example, the system can only retain valid history (such as records whose quality of service exceeds a certain minimum threshold) to save storage space.

Step 303: When the user applies for the relay service, the relay control system selects the relay node according to the valid relay service history record.

Specifically, if there is a valid relay service history record, and the source address/group ID and the destination address/group ID are respectively associated with the source address/group ID and the destination address/group involved in the user requesting the relay service. If the group ID is the same, the relay control system can directly use the relay described in the relay service history record as the candidate relay node.

It should be noted that, in step 301, the relay service information collection may take multiple manners. FIG. 4 is a schematic flowchart of a method for collecting service quality information according to an embodiment of the present invention. In this embodiment, the service quality information collection is the same. After the service is carried out simultaneously, when the relay node A is the first, When the second network node provides the relay service, the service quality of the current service is measured, and the service quality information is reported to the relay control node after the service ends. As shown in FIG. 4, the method specifically includes:

Step 401: The first network node and the second network node establish a data link through the relay node A to perform a certain service.

Step 402: The relay node A counts the service quality of the service during the service process, and records the service.

Here, the quality of service statistics may be completed by the relay node A or by the first network node or the second network node. The statistical record can be aggregated to the relay node A, or the first network node or the second network node.

Step 403: After the service is completed, the relay node A reports the recorded quality of service information of the service to the relay control node.

Here, the service quality information may also be reported to the relay control node by the first network node or the second network node, and the service quality information report may be reported to the relay control node after the service is completed or during the service progress, and the quality of service record may be once. It can also be reported in multiple times.

It should be noted that, in addition to the quality of service, in the process of collecting and reporting the service quality information, other relay service information (that is, other parameters that should be included in the relay service history record) may also be collected and reported.

FIG. 5 is a schematic flowchart of another method for collecting service quality information according to an embodiment of the present invention. In this embodiment, a relay control node actively requests a relay node A to perform service quality measurement. As shown in FIG. 5, the method includes:

Step 501: The relay control node requests the relay node A for its quality of service history from the first network node (source node) to the second network node (destination node).

Here, in addition to the quality of service, the relay control node may also request relay node information (ie, other parameters that should be included in the relay service history) to the relay node A at the same time. Step 502: The relay node A does not have an associated quality of service history, and then establishes a data link to the first network node and the second network node to perform service quality measurement.

Before this step, optionally, the relay control node may notify the first network node and the second network node to prepare a quality of service measurement. Optionally, the relay control node may also establish a link of the first network node to the second network node in the relay node A through the control channel.

Step 503: The relay node A sends the obtained quality of service history to the relay control node.

Here, the quality of service information may also be reported by the first network node or the second network node. FIG. 6 is a schematic flowchart of a method for collecting service quality information according to an embodiment of the present invention. In this embodiment, a relay control node requests a service quality record from another network information server (which may be a physical entity or a logical entity) through other interfaces. As shown in FIG. 6, the method includes:

Step 601: The relay control node needs the service quality history information between the certain groups, and the relay control node determines that the network information server A can provide related information (the determination process can be through DNS query, database query or static configuration, etc.) In a plurality of manners, the relay control node fills in the service quality information request and sends it to the network information server A according to the range of the required information.

Here, in addition to the quality of service, the relay control node may also obtain other relay service information (ie, other parameters that should be included in the relay service history) to other network information servers.

Step 602: The network information server A obtains corresponding network service quality information. The network information server A may choose to initiate the quality of service measurement or query the own database to obtain the relevant information. Step 603: The network information server A returns the quality of service record information according to the requirements of the relay control node.

Network Information Server A can be other relay control nodes, routers or other network devices that can obtain quality of service information.

The process shown in Figure 6 describes that the relay control node actively requests the quality of service history from other nodes. The process of recording, in actual applications, other nodes can also actively push the quality of service history to the required relay control node. Of course, other nodes (network information servers) can also push other relay service information to the required ones at the same time. Following the control node.

FIG. 7 is a schematic flowchart of relay selection using a relay service history record according to an embodiment of the present invention. As shown in FIG. 7, the method includes:

Step 701: The first network node (source node) needs to use the relay service to communicate with the second network node (the destination node), and the first network node sends the address of the communication party to the relay control node by using the relay service request. Request a relay resource.

Here, the relay service request may be initiated by the first or second network node, or may be forwarded by other service entities.

Step 702: After receiving the address information of the first network node and the second network node, the relay control node queries the history database, and the query may be based on the address, network segment, group ID, service type, or access of the source destination node. The method queries for the query; the database query returns one or more valid relay service history records, and the relay control node determines that the relay (relay node A) used in the relay service history record can carry the current service (for example, its source) The destination group is consistent with the new application).

Here, if the relay control node does not query the valid relay service history record, the relay control node may use the original relay selection algorithm to determine the candidate group based on the inter-group measurement result and perform further measurement.

Step 703: The relay control node interacts with the selected relay node A to reserve a relay resource.

In step 702 and step 703, the relay control node may also select a plurality of qualified relay nodes at the same time and perform resource reservation.

Step 704: The relay control node returns the information of the relay node A to the first network node. Step 705: The first network node and the second network node establish communication through the relay node A. It can be seen that the use of history in the relay selection process can reduce measurement information and selection. Choose algorithm delay.

It should be noted that the relay control system may also adjust the group classification information (the correspondence between the group and the network address) according to the relay service history record, and the nodes in the group use the same relay node but obtain a very different difference. When the quality of service is available, the system can automatically subdivide the group into groups. When nodes in different groups use the same relay node, they will always get the same quality of service. These groups can be combined. The specific implementation of group segmentation and group combination is as follows:

Group subdivision: Suppose there are multiple history records with approximate timestamps, and there are two addresses in the first second network node and the relay node address triplet. The third address is inconsistent but in the same group. in. For example, there are multiple history records that have the same relay node address, a second network node address, and the first network node has the same group ID (but the address is different). If the quality of service information at this time indicates that the quality of service difference of the nodes is not less than a certain threshold (preset value), the system will try to subdivide the group to which the first network node belongs, that is, the first network node with the same quality of service. Address (or access method) categorizes the original group into two or more subgroups. If the classification fails, the system can choose to mark the group non-use history for selection, or choose not to process.

Group merging: If the system analyzes two groups, the service quality of the related history is almost the same under similar conditions, that is, the difference of the QoS is not greater than the preset value, the system can merge the two groups. . For example, suppose that group A and group B each have multiple history records, and the system checks that all of the first second network node and the relay node address group ID triplet are identical, and the third is a group. A or group B records, if the quality of service provided by these records is basically the same, and the number of records exceeds a certain limit, the system can consider that group A and group B are equivalent, and merge group A and group B.

8 is a schematic structural diagram of a relay selection system according to the present invention. As shown in FIG. 8, the system includes: a source node, a relay node, a destination node, and a relay control system. When the source node wants to communicate with the destination node, if the network Quality problem or source/destination node is in NAT/firewall After that, direct communication cannot meet the needs of users. Then the source node or the destination node can request communication through one or more relay nodes. The relay node is responsible for the data information between the destination nodes of the transit source. At the same time, the relay node is controlled by the relay control system. The relay control system accepts the relay request of the source destination node, and at the same time, the relay control system is responsible for selecting an appropriate relay node and maintaining the network measurement information necessary for the relay node selection algorithm.

Specifically, the relay control system further includes: a relay service information collection module, a relay service history record formation module, a relay service history record processing module, and a relay node selection module;

The relay service information collection module is configured to collect relay service information;

The relay control system requests relay service information from other network information servers, or the relay control system acquires relay service information actively pushed by other network information servers. The relay service history forming module forms the relay service information as: storing the relay service information in the server, the related disk array or the network storage, or the relay service letter The information is stored in a distributed storage manner.

Delete expired history based on the timestamp of the history;

Sort and aggregate historical records;

Keep records of the best quality of service in the same type of history.

In actual implementation, the relay control system can be implemented by a single server or by multiple physical entities. The relay node and the source destination node (source node and destination node) can also be combined with other network entities. The present invention only describes the interaction between logical entities, and its signaling flow and implementation can be used for a variety of different network architectures.

The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the present invention.

Claims

Claim

A relay selection method, wherein the method comprises:

The relay control system processes the relay service history record to obtain a valid relay service history record;

2. The method according to claim 1, wherein the relay service history record comprises one or more of the following: a relay node, a source node, and an address/group ID of the destination node involved in the relay service; Service quality information following the service; timestamp of the relay service.

3. The method according to claim 2, wherein the relay service history record further comprises one or more of the following: a relay node, a source node, and an access mode of the destination node involved in the relay service; Business type of business.

4. The method according to claim 1, wherein the quality of service is a quality of service between a source node and a destination node, or a quality of service between a source node/destination node and a relay node, the quality of service It is manifested by one or more of the following: delay, delay jitter, bandwidth, packet loss rate.

5. The method according to claim 1, wherein the relay control system collects relay service information as:

The relay control system requests relay service information from other network information servers, or The relay control system acquires relay service information actively pushed by other network information servers.

6. The method according to claim 1, wherein the relay control system saves the relay service information as: the relay control node in the relay control system is a single centralized server, and the relay service information is stored in On the server, in the relevant disk array or network storage; in the relay control system, there are multiple peer-to-peer relay control nodes, and the relay service information is distributed on different relay control nodes in a distributed storage manner.

7. The method according to claim 2, wherein the relay control system processes the relay service history record, and obtaining a valid relay service history record comprises one or more of the following:

Delete expired history based on the timestamp of the history;

Sort and aggregate historical records;

Keep records of the best quality of service in the same type of history.

8. The method according to claim 2, wherein, when the user applies for a relay service, the relay control system selects the relay node according to the valid relay service history record as:

The method according to any one of claims 1 to 8, wherein the method further comprises: the relay control system adjusting the group classification information according to the relay service history, specifically:

10. A relay control system, wherein the system comprises: a relay service information collection module, a relay service history record formation module, a relay service history record processing module, and a relay node selection. Module; among them,

11. The system according to claim 10, wherein the relay service information collection module collects relay service information as:

The system according to claim 10, wherein the relay service history forming module forms the relay service information as: storing the relay service information in the server, the related disk array or the network storage, or The relay service information is stored in a distributed storage manner.

13. The system according to claim 10, wherein the relay service history processing module processes the relay service history record, and the valid relay service history record includes one or more of the following:

Delete expired history based on the timestamp of the history;

Sort and aggregate historical records; Keep records of the best quality of service in the same type of history.

The system according to claim 10, wherein the relay node selection module selects a valid relay service history record after processing by the relay service history processing module when the user applies for the relay service. Following the node is:

The system according to any one of claims 10 to 14, wherein the system further comprises a group classification information adjustment module, configured to adjust the group classification information according to the relay service history record, specifically: