WO2009132545A1

WO2009132545A1 - Root node and method for constructing multi-tree topology

Info

Publication number: WO2009132545A1
Application number: PCT/CN2009/071196
Authority: WO
Inventors: 韩磊; 梁超; 罗斯·基思; 刘勇; 李峰
Original assignee: 华为技术有限公司
Priority date: 2008-04-29
Filing date: 2009-04-08
Publication date: 2009-11-05
Also published as: CN101572620B; CN101572620A

Abstract

The present invention discloses a root node and a method for constructing a multi-tree topology in the network management field. The method includes the following steps: determining the role of node according to uplink bandwidth information, determining the sub-tree the node is attached to according to the principle of sub-trees’ uplink bandwidths equality or approximate equality; determining the location of the node in the sub-tree the node is attached to according to the uplink bandwidth information of the node. The root node includes a module for determining the role, a module for determining the sub-tree the node is attached to and a module for determining the location. The present invention considers the uplink bandwidth information of node when constructing a multi-tree topology, allocates nodes to each sub-tree evenly according to uplink bandwidth information of the node, guarantees the convenience of the sub-tree depth control in multi-tree topology and improves transmission efficiency of multi-tree topology.

Description

A method for constructing multi-tree topology and a root node

This application claims priority to Chinese Patent Application No. 200810105449.9, entitled "A Method for Constructing a Multi-Tree Topology and a Root Node", filed on April 29, 2008, the entire contents of which are hereby incorporated by reference. The citations are incorporated herein by reference. Technical field

The present invention relates to the field of network management, and in particular, to a method for constructing a multi-tree topology and a root node. Background technique

With the continuous development of broadband multimedia networks, various broadband network applications are emerging. Broadband applications such as Internet Protocol Television (IPTV), video conferencing, data and data distribution, network audio applications, network video applications, and multimedia distance education pose challenges to the carrying capacity of existing broadband multimedia networks. Traditional networks built with unicast technology are no longer able to meet the bandwidth and network service quality requirements of emerging broadband network applications, which are followed by network delays and data loss. This introduces IP multicast technology to solve the above problems. IP multicast technology is an important extension of the basic "unicast, best-effort" model of the existing Internet. It implements a point-to-multipoint network connection between the sender and each receiver. If a sender transmits the same data to multiple recipients at the same time, it only needs to copy the same data. IP multicast technology improves data transfer efficiency and reduces the possibility of network congestion. At present, a lot of research has been done on IP multicast technology. However, IP multicast technology is not widely used in the Internet because of serious problems in transmission technology and management.

In recent years, with the introduction of technologies such as Peer-to-Peer Network (P2P network) and Overlay Network (Laser Network), Application Layer Multicast (ALM) technology has emerged. Its main idea is: to maintain the original "unicast, best-effort" function of the Internet, try not to change the original network architecture, and mainly achieve the multicast function by adding the functions of the end system. Application layer multicast only needs to change the end system without any modification to the router, which is easy to implement and promote. In addition, application layer multicast can design different implementation solutions for different applications, and it is flexible. It is not necessary to unify different applications into one model like IP multicast technology.

Multi-tree topology-based application layer multicast, which maintains multiple multicast trees simultaneously between multicast members. The system encodes the media data into N (N is a positive integer) substreams using a Multiple Description Coding (MDC) algorithm. Wherein, each substream has the same bandwidth. N substreams are simultaneously propagated along multiple multicast trees. The multicast member can complete decoding as long as it receives any M (M is a positive integer) substream of the N substreams, where M<N. The application layer multicast scheme based on multi-tree topology greatly improves the robustness of the system, and reduces the heterogeneity of the terminal and the influence of the random number of incoming and outgoing terminals on the transmission effect of the system. Therefore, how to build a multi-tree topology is an important issue when implementing an application layer multicast scheme.

There are many methods for constructing a multi-tree topology in the prior art. A representative method is the Scribe method. The characteristics of the multi-tree topology constructed according to the Scribe method are as follows: First, if the node is in a subtree As an intermediate node, it can only serve as a leaf node in other subtrees. Second, using the node ID (identification) as input, and using the Pastry algorithm to obtain the specific location of the node. Specifically, five nodes in the system are node A, node B, node C, node D, and node E, and the uplink bandwidth is 256K bits, 512K bits, 1M bits, 2M bits, and 4M bits, respectively, and the current single substream needs The upstream bandwidth is 512K bits, and the multi-tree topology constructed according to the Scribe method is shown in Figure 1. In the figure, the first subtree of the multi-tree topology has a depth of 4, and the second subtree has a depth of 2. Since the overall delay of the multi-tree topology is affected by the depth of the largest subtree, the deeper the subtree depth, the larger the delay, and the greater the overall delay of the multitree topology.

The inventors found during the research that the existing method of constructing a multi-tree topology would result in the subtree depth of the multi-tree topology being difficult to control. Summary of the invention

In order to make the subtree depth of the multi-tree topology easy to control, the embodiment of the present invention provides a method for constructing a multi-tree topology and a root node. The technical solution is as follows:

A method of constructing a multi-tree topology, the method comprising:

Determining a role of the node according to the uplink bandwidth information of the node, and determining a subtree to which the node belongs according to a principle that the uplink bandwidth of the subtree is equal or similar;

Determining, according to the uplink bandwidth information of the node, a location of the node in the belonging subtree.

A root node, the root node comprising:

a role determining module, configured to determine a role of the node according to uplink bandwidth information of the node;

a subtree attribution determining module, configured to determine, according to a principle that the uplink bandwidth of the subtree is equal or similar, the subtree to which the node belongs;

And a location determining module, configured to determine, according to the uplink bandwidth information of the node, a location of the node in the subtree to which it belongs.

In the embodiment of the present invention, when the multi-tree topology is constructed, the uplink bandwidth information of the node is considered, and the nodes are allocated to each subtree according to the uplink bandwidth information of the node, thereby ensuring that the subtree depth of the multi-tree topology is easily controlled, thereby facilitating Improve the transmission efficiency of multi-tree topology. DRAWINGS

1 is a schematic diagram of a multi-tree topology constructed based on the Scribe method provided by the prior art;

2 is a flow stalk diagram of a method for constructing a multi-tree topology according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of determining a hierarchy of subtrees to which a node belongs according to an embodiment of the present invention;

4 and FIG. 5 are another schematic diagram of determining a hierarchy of subtrees to which a node belongs according to an embodiment of the present invention; FIG. 6 is a schematic diagram of multi-tree topology adjustment when a node exits according to an embodiment of the present invention;

FIG. 7 is a schematic diagram of a multi-tree topology after completion of construction according to an embodiment of the present invention; FIG.

FIG. 8 is a schematic structural diagram of a root node according to an embodiment of the present invention. detailed description

The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

The embodiment of the invention provides a method for constructing a multi-tree topology, which is based on the uplink bandwidth information of the node to construct a multi-tree topology, so that the depth of each sub-tree of the multi-tree topology is easily controlled, thereby helping to improve the multi-tree topology. Transmission efficiency. As shown in FIG. 2, the specific steps of this embodiment are as follows:

101: Determine a role of the node according to the uplink bandwidth information of the node, and determine a subtree to which the node belongs according to the principle that the uplink bandwidth of the subtree is equal or similar.

The specific method is: the root node compares the uplink bandwidth of the node with the sub-stream bandwidth of the sub-tree. If the uplink bandwidth of the node is smaller than the sub-stream bandwidth of the sub-tree, the node is added as a leaf node to all sub-trees; The uplink bandwidth is greater than the sub-stream bandwidth of the sub-tree, and the node is added as an intermediate node to the sub-tree with the smallest uplink bandwidth, and the node is used as a leaf node in other sub-trees. The uplink bandwidth of the subtree is equal to the sum of the uplink bandwidth of the intermediate nodes. The greater the sum of the uplink bandwidths of the intermediate nodes, the greater the uplink capability of the subtree.

In the embodiment of the present invention, in order to ensure that the node failure has the least impact on the system, if a node acts as an intermediate node in a subtree, it can only act as a leaf node in other subtrees.

102: Determine the level of the node in the subtree to which it belongs.

If the role of the node is a leaf node, the node is at the lowest level of the subtree; if the role of the node is an intermediate node, the uplink bandwidth information of the node is sent to the root node, and the root node has the uplink bandwidth of the node and each node. The uplink bandwidth is compared. If the uplink bandwidth of the node is smaller than the maximum uplink bandwidth of the node in the current layer and greater than the minimum uplink bandwidth of the node in the current layer, it is determined that the node belongs to the current layer and joins the current layer. If the current layer still has a storage space, the node directly joins the current layer. If the current layer has insufficient storage space, it replaces the upstream bandwidth in the current layer. The smallest node joins the current layer. As shown in FIG. 3, the node G joins the subtree as an intermediate node, and the node G sends its uplink bandwidth information to the root node. After the root node compares the uplink bandwidth of the node G with the uplink bandwidth of each layer node, it is found that the uplink bandwidth of the node G is smaller than the maximum uplink bandwidth of the node in the third layer, which is greater than the minimum uplink bandwidth of the node in the third layer. It is determined that the node G belongs to the third layer. Since the fan-out capability of the second-layer node is 6, it indicates that the number of nodes that can be served by the second-layer node is 6, and the third layer has two nodes, indicating that the third layer still has a storage space. G directly joins the third layer. As shown in FIG. 4 and FIG. 5, the node H joins the subtree as an intermediate node, and the node H sends its uplink bandwidth information to the root node. The root node compares the uplink bandwidth of the node H with the uplink bandwidth of each layer node, and finds that the uplink bandwidth of the node H is smaller than the maximum uplink bandwidth of the node in the third layer, which is greater than the minimum uplink bandwidth of the node in the third layer. Therefore, It is determined that the node H belongs to the third layer. Since the fan-out capability of the third layer is 3, the number of nodes that can provide services to the second-layer node is 3, and the third layer already has 3 nodes, indicating that the storage space of the third layer is insufficient. H squeezes the node H with the smallest upstream bandwidth in the third layer to the fourth layer, occupying the position of the original node H.

103: Determine the parent node of the node in the subtree to which it belongs.

After determining the level in the subtree to which the node belongs, the node generates a test message carrying its own capability information, and sends the test message to all nodes of the upper layer, and the upper layer node adds its capability information to the receiving. In the test message, send the test message to its upper-level node, and so on, until the test message reaches the root node. As can be understood by those skilled in the art, when a node sends its capability information to an upper node through a test message layer by layer, the upper node also adds its capability information to the test message and sends it to its upper node until the test message reaches Root node. The root node determines the optimal path of the system according to the preset selection policy, and returns the best path to the node, and the node accesses the best path selected by the root node. It should be noted that the node's own capability information can be selected according to requirements. For example, each node sends a test message carrying its own available uplink bandwidth information to the root node, and the root node receives the available uplink bandwidth information of each node, according to the path. The principle is to return the node on the path with the largest upstream bandwidth to the requesting node as its parent.

After the multi-tree topology is constructed, the root node records the location information and uplink bandwidth information of each layer node.

104: Adjust the multi-tree topology when a node exits.

In practical applications, the node will often exit the system. In order to reduce the impact of node exit on the system, the exit mechanism is:

The child node of the exit node directly carries its descendant node to perform the access process, and the descendant node does not change its parent-child relationship, so that the positional relationship change of the node is small, and the topology oscillation caused by the node exit is greatly reduced. As shown in FIG. 6, it is assumed that the uplink bandwidth of the node A is greater than the uplink bandwidth of the node B. When the grandfather node G of the node B exits, the child node F of the node G carries its child node B to perform the rejoining process, and the positional relationship between the node F and the node B does not change. Chemical. Since the uplink bandwidth of the node A is greater than the uplink bandwidth of the node B, the positional relationship between the node A and the node B is adjusted. Since the grandfather node exits, the positional relationship between its child node F and its descendant nodes does not change. Therefore, it helps to reduce the topological oscillation caused by node exit.

After the node exit adjustment is completed, the root node records the location information of the nodes in the subtree and the uplink bandwidth information of each layer node. Specifically, it is assumed that the system has nine nodes, namely node A, node B, node C, node D, node E, node F, node G, node H, and node I, and the uplink bandwidth of each node is 256K, 256Κ, respectively. 2Μ, 4Μ, 4Μ, 256Κ, 256Κ, 2Μ, 1Μ, the sub-stream bandwidth of the subtree is 1Μ, the fan-out capability of the root node in each subtree is 2, and the number of subtrees preset by the user is 2. The multi-tree topology constructed according to the technical solution of this embodiment is shown in FIG. 7, in which: node Α, node Β, node F, and node G are leaf nodes, and node C, node D, node E, node H, and node I may As an intermediate node. The intermediate node in the first subtree. The sum of the uplink bandwidths of E, I and E is 256K+1M+4M, and the sum of the uplink bandwidths of the intermediate nodes D and H in the second subtree is 4M+2M, so that the uplink bandwidths of the two subtrees are approximate, and the depths of the two subtrees are Both are 2, so the transmission efficiency of the multi-tree topology is improved.

In the embodiment of the present invention, when constructing a multi-tree topology, the uplink bandwidth information of the node is considered, and the nodes are allocated to each subtree according to the uplink bandwidth information of the node, thereby ensuring that the subtree depth of the multi-tree topology is easy to control, and thus Helps provide transmission efficiency for multi-tree topology. An embodiment of the present invention provides a root node. As shown in FIG. 8, the root node includes:

a role determining module 800, configured to determine a role of the node according to uplink bandwidth information of the node;

The subtree attribution determining module 810 is configured to determine, according to the principle that the uplink bandwidth of the subtree is equal or similar, the subtree to which the node belongs;

The location determining module 820 is configured to determine, according to the uplink bandwidth information of the node, the location of the node in the subtree to which it belongs.

The location determining module 820 includes:

a hierarchy determining module 821, configured to determine a level of the node in the belonging subtree;

The parent node determining module 822 is configured to determine the parent node of the node in the belonging subtree.

Further, the root node further includes:

a triggering module 830, configured to trigger a role determining module 800, a subtree attribution determining module 810, and a location determining module 820 when a node exits;

The role determination module 800 determines the role of the child node of the exit node in the subtree; the subtree attribution determination module 810 determines the subtree to which the child node of the exit node belongs; the location determination module 820 determines that the child node of the exit node is in the belonging child The location in the tree. The detailed functions of the role determination module, the subtree attribution determination module, and the location determination module may be referred to as 101, 102, and 103, respectively, and details are not described herein.

In the embodiment of the present invention, when constructing a multi-tree topology, the uplink bandwidth information of the node is considered, and the nodes are allocated to each subtree according to the uplink bandwidth information of the node, thereby ensuring that the subtree depth of the multi-tree topology is easily controlled. , which in turn helps to improve the transmission efficiency of the multi-tree topology.

The technical solution provided by the above embodiments can be implemented by hardware and software, and the software is stored on a readable storage medium, such as a floppy disk, a hard disk or an optical disk of a computer.

The above is only the preferred embodiment of the present invention, and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., which are within the spirit and scope of the present invention, should be included in the protection of the present invention. Within the scope.

Claims

Claim

A method for constructing a multi-tree topology, the method comprising:

The method for constructing a multi-tree topology according to claim 1, wherein the determining the role of the current node according to the uplink bandwidth information of the node, and determining the node according to the principle that the uplink bandwidth of the sub-tree is equal or similar The steps of the belonging subtree specifically include:

Comparing the uplink bandwidth of the node with the sub-stream bandwidth of the sub-tree, if the uplink bandwidth of the node is greater than the sub-stream bandwidth of the sub-tree, adding the node as an intermediate node to the sub-tree with the smallest uplink bandwidth, in other As a leaf node in the subtree, if the uplink bandwidth of the node is smaller than the substream bandwidth of the subtree, the node is added as a leaf node to all subtrees.

The method for constructing a multi-tree topology according to claim 1, wherein the step of determining the location of the node in the subtree to which the node belongs according to the uplink bandwidth information of the node includes:

Determining a hierarchy of the nodes in the assigned subtree;

Determining the parent node of the node in the belonging subtree.

The method for constructing a multi-tree topology according to claim 3, wherein the step of determining the level of the node in the classified subtree comprises:

Comparing the uplink bandwidth of the node with the uplink bandwidth of the current layer node, if the uplink bandwidth of the node is smaller than the maximum uplink bandwidth of the node in the current layer, and greater than the minimum uplink bandwidth of the node in the current layer, determining the node Belongs to the current layer.

The method for constructing a multi-tree topology according to claim 3, wherein the step of determining the parent node of the node in the belonging subtree comprises:

Receiving the test information sent by the node and the upper node of the node and carrying the self, and selecting a parent node for the node according to a preset policy.

6. A method of constructing a multi-tree topology according to any of claims 1-5, wherein: The method also includes:

When a node exits, the child node of the exit node carries its child node to perform the join process.

7. A root node, wherein the root node comprises:

a role determining module (800), configured to determine a role of the node according to uplink bandwidth information of the node; a subtree attribution determining module (810), configured to determine, according to a principle that the uplink bandwidth of the subtree is equal or similar, the node belongs to Subtree

The location determining module (820) is configured to determine, according to the uplink bandwidth information of the node, a location of the node in the belonging subtree.

The root node according to claim 7, wherein the location determining module (820) comprises: a level determining module (821), configured to determine a level of the node in the belonging subtree;

The parent node determination module (822) is configured to determine the parent node of the node in the assigned subtree.

The root node according to claim 7 or 8, wherein the root node further comprises:

a triggering module (830), configured to trigger the role determining module (800), the subtree attribution determining module (810), and the location determining module (820) when a node exits;

The role determining module (800) determines a role of the child node of the exiting node in the subtree; the subtree attribution determining module (810) determines a subtree to which the child node of the exiting node belongs; the location determining module (820) determining a location of the child node of the exit node in the belonging subtree.