CN1822548A - Distributed traffic management method based on peer-to-peer network - Google Patents

Abstract

Present invention relates to distributing type flow rate management method Based on peer to peer network, utilizing third generation P 2 P structure through all user installed client-side and combined with special server in network, to realize intra-network user visiting flow rate management and control. Said method functional entity includes common node, super node and authenticating server, catalog server, strategy server and information server. Said method adopts feed-back system, according to load status to set flow rate management strategy with high efficiency and accuracy flow rate identification, reducing flow rate management equipment load and cost.

Description

Distributed traffic management method based on peer-to-peer network

technical field

The present invention is a method for realizing the management and control of network traffic, adopts peer-to-peer network to realize the management and control of network traffic, relates to the application of peer-to-peer network architecture, and belongs to the field of network traffic management

technology field.

Background technique

With the continuous development of current network technology, new network services emerge in an endless stream, among which P2P (peer-to-peer) network sharing services and streaming media services are particularly prominent. The emergence of new network services has brought a huge impact on traditional Internet applications and brought challenges to network operation and maintenance management. In particular, the bandwidth preemption characteristic of the P2P network mode brings great challenges to network resource management. According to statistics, at present, domestic and foreign P2P traffic accounts for more than 60% of the total network traffic, which seriously affects the operation of some normal services and causes a waste of network bandwidth resources. Therefore, with the emergence of new network services, the research on network traffic identification and management technology and the design of network traffic management system have become the current research hotspots.

The research and development of P2P network technology has promoted the wide application of P2P technology, including traditional file sharing, the research of P2P-based SIP protocol, the use of P2P to realize IP telephony, and the research of P2P-based distributed database system.

At present, most systems for traffic identification and management deploy traffic management devices centrally and serially at the network egress, realize service identification for data flowing through the devices, and divide, rate limit, and discard data according to different traffic management strategies. Some similar flow pipe devices adopt a distributed architecture, but the so-called distributed architecture is nothing more than deploying multiple flow pipe devices on multiple internal backbone branches of the network to implement traffic management for each branch.

In the prior art, a centralized traffic management device is mainly used, which is deployed at a network backbone branch or a network egress to manage and control data flowing through it. The disadvantages of the prior art are:

1) It is difficult to achieve accurate traffic identification. In the existing technology, network branches or outlets can only identify the passing data according to the characteristics of the data. However, with the development of new network services, the network traffic identification technology is becoming more and more complex, and it is difficult to achieve accurate identification only based on the characteristics of the data. traffic identification.

2) The traffic management equipment has heavy load and high cost. For the traffic management device deployed on the network backbone or network exit, all incoming and outgoing data in the network must flow through the device, be identified by the identification strategy, and be controlled by the control strategy, so the load on the device is heavy Heavy and high performance requirements, so as the maximum flow allowed by the equipment increases, the cost of the equipment will also increase dramatically.

3) Intra-network traffic management cannot be realized. Devices deployed at the network egress or network backbone can only manage the traffic flow of the passing data, so they cannot implement intra-network traffic management. In this way, although the management and control of network traffic can be realized at the network egress, the management and control of data transmission between users in the network cannot be realized.

Contents of the invention

Technical problem: the purpose of this invention is to propose a distributed traffic management method based on peer-to-peer network, that is, to adopt the current popular P2P technology to realize the method of distributed network traffic management, and to adopt the feedback mechanism to effectively realize the identification and management of network traffic. Management, especially the effective identification and management of new network services such as streaming media services and P2P.

Technical solution: The present invention proposes a method for implementing distributed traffic management using the P2P principle, which introduces the overall structure and implementation principle of the method, proposes each functional entity included in the distributed traffic management method based on P2P, and introduces each The interaction mechanism of functional entities, especially the feedback mechanism is applied in this method. The method adopts the third-generation P2P architecture, that is, hybrid P2P implementation, which is similar to the third-generation P2P systems such as Skype, BT, etc. This method includes three types of functional entities: ordinary nodes, super nodes and special servers. Different entities have different configurations. In the method In the method, ordinary nodes and super nodes are both ordinary network users. As a network access pass, the corresponding functional software is installed on the computers of ordinary network users. The software is determined according to the configuration of the user's computer. It is also the function of a super node; the special server in the method is one or more servers with different functions configured in the network, including a traffic management policy server, a network information server, and the like. In addition, the method adopts a feedback mechanism, and traffic management strategies are different according to different network load conditions.

The following introduces the overall architecture and working principle of the P2P-based distributed traffic management method, each functional entity and the interaction of the entities.

Firstly, the overall architecture and working principle of the P2P-based distributed traffic management method are introduced.

The method adopts the third-generation hybrid P2P network architecture, which has a functional entity similar to Skype, and has a three-layer functional structure. From bottom to top, there are ordinary node layer, super node layer and server layer, which correspond to three types of functional entities.

All users in the network must install software with local traffic identification and management functions with the identity of the network ID card to access the network. That is, if each user in the network needs to access the Internet, he must first install the software, similar to Dr.COM broadband login Client, and according to the performance of the user's computer, the software automatically determines whether the user belongs to a normal node or a super node. For ordinary nodes and super nodes, the software has basic functions such as network access, traffic identification, and traffic management. information and submit it to the information server and other special functions.

In order to provide effective management for ordinary nodes and super nodes, the method contains a number of special servers. These servers do not manage network traffic like traditional traffic management equipment, nor are they the outlet of network data. They are only used to manage common Nodes and super nodes, and provide necessary information for super nodes or other servers, such as directory server provides super node list information for other special servers, policy server provides traffic management policies for super nodes, information server summarizes network load provided by super nodes, etc. information and submit it to the policy server.

The functional entities and entity interactions contained in the method are introduced in detail below.

1. Ordinary nodes:

Ordinary nodes are network users who have installed client software but have relatively poor computer performance. It has the most basic functions of client software. The specific working process includes:

1) User authentication for local access to the network. Only by using the client for user information authentication can the user have access to the network. The newly added client connects to the authentication server for user authentication, and then the authentication server communicates the user information to the common node The super node to which it belongs is managed by the super node;

2) After obtaining the network access user authentication, the directory server obtains the locally accessible supernode list information, and establishes a connection with the supernode;

3) Obtain the network access and data transmission status of all local active processes, and submit the local network transmission status to the super node;

4) Effectively manage and control the network access and data transmission of all local processes according to the traffic management strategy conveyed by the super node, and realize traffic shaping.

2. Super nodes: Super nodes are network users who have installed client software and have relatively good computer performance. In addition to the basic functions of ordinary node clients, it also has the functions of managing ordinary nodes, communicating network transmission information in sub-domains, and communicating traffic management policies, so as to finally achieve the purpose of slowing down the load of special servers in the center.

Compared with ordinary nodes, super nodes have the following characteristics;

1) User authentication for accessing the network is directly connected to the authentication server, and the newly added super node information is communicated to the directory server by the authentication server;

2) When the directory server communicates the super node information to the newly added common child node, it transmits the basic information of the child node to the super node;

3) Summarize the transmission information of each common node, and convey the network transmission information within the jurisdiction to the information server;

4) The traffic management policy is obtained by the policy server, and communicated to the common nodes in the domain requesting the policy.

3. Special server:

Special servers are special entities within the network, similar to login servers and event servers in Skype. According to actual needs, the special server designed by this method includes a total of four special servers: authentication server, directory server, policy server, and information server. The specific functions are as follows:

1) Authentication server:

Realize the function of client user access network authority authentication, store all client user authentication information on the server, and have functions such as new user registration and password modification. The specific information interaction includes: receiving the registration information of the newly added node, receiving the password modification information of all clients, and conveying the information of the newly added node to the directory server, etc.

2) Directory server:

The directory server saves all super node information, can assign the list of super nodes to newly added ordinary nodes, can communicate the ordinary node information of the requested super node to the corresponding assigned super node, and communicate the list of all super node information to the policy server and The information server facilitates the policy server to distribute policies and the information server to summarize network transmission information.

3) Policy server:

According to the overall transmission status of the network obtained by the information server, including the load status of each part of the network, the policy server formulates traffic management and control policies based on load feedback for different policy request points, and communicates the policy to the super node requesting the policy. In turn, the super node communicates to the ordinary node requesting the policy.

4) Information server:

The main function of the information server is to receive the network transmission information conveyed by the super nodes, summarize and analyze all the network transmission information within the jurisdiction of each super node, and make statistics, and pass the statistical results of the network transmission information to the policy server, so that the policy server can follow the According to the current load status of the network, a traffic management strategy based on load feedback can be formulated, so that each client can assign different permissions and traffic shaping to the local process access network according to the current network transmission status, and finally achieve full and reasonable use of network resources the goal of.

The distributed traffic management method based on the peer-to-peer network of the present invention adopts the third-generation peer-to-peer network architecture, that is, adopts the peer-to-peer network principle to realize distributed network traffic identification and management at each network client; the method functional entities include common nodes , super node and authentication server, directory server, policy server and information server; in addition, this method uses a feedback mechanism to formulate a traffic management strategy according to the network load status. The specific traffic management method is as follows:

1) Users in the network install the client software, generate a list of special servers, and judge the configuration status of the local computer when running for the first time after installation, so as to determine whether it is a normal node or a super node;

2) The client software runs and communicates local basic information and new user registration information to the authentication server. If it is a common node, the directory server obtains the list of super nodes it belongs to, and the directory server communicates the common node information to the super node it belongs to. , and finally establish a connection with the super node; if it is a super node, the authentication server will communicate the super node information to the directory server to update the super node list;

3) When ordinary nodes have network access and data transmission behaviors, they report the local network status to super nodes, and obtain traffic management policies through super nodes;

4) The super node communicates the network transmission information of the ordinary nodes under its jurisdiction to the information server, and the information server feeds back to the policy server, and the policy server communicates the traffic management policy to the super node, and the super node communicates it to the ordinary node requesting the policy;

5) The authentication server updates the authentication information of all users on the network in a timely manner, and constantly exchanges information with the directory server to update the super node list information stored in the directory server;

6) The information server receives the network transmission information conveyed by each super node in a timely manner, and feeds back the network transmission status to the policy server, so as to formulate a traffic management policy with a feedback mechanism, ensure the rationality of policy formulation, and fully and rationally use network resources.

The overall architecture to implement this method is:

1) Adopt the third-generation hybrid P2P network architecture, with a functional entity similar to Skype;

2) It has a three-layer functional structure, which is the ordinary node layer, super node layer and server layer from bottom to top, corresponding to three types of functional entities in turn;

3) Network access through all network users must be realized by installing software with local traffic identification and management functions as a network access ID card;

4) Adopt a feedback mechanism to formulate traffic management strategies according to the overall transmission status of the network.

The execution process of the client software installed by ordinary nodes and super nodes is:

1) The client software runs first to determine whether it is running for the first time, and if so, checks the configuration status of the local computer to determine whether the user can become a super node,

a. If it meets the super node, register the super node with the authentication server, and the authentication server will notify the directory server to update the list of super nodes, and then perform the same process as the non-first run of the super node;

b. If it does not qualify as a super node, register a normal node with the authentication server, and connect to the directory server to obtain a list of the nearest super nodes, and then perform the same process as the normal node for non-first run;

2) If the client software is not running for the first time, if it is a normal node, then:

c. Detect the super node list and establish a connection with a certain super node;

d. Pass the user authentication of the super node and obtain network access rights;

e. Monitor the current local network transmission status and report to the super node;

f. Receive policy information delivered by super nodes to realize local traffic management and control;

3) If the client software is not running for the first time, if it is a super node, then:

g. Pass the user authentication of the authentication server and obtain network access rights;

h. Listen to and receive connections from common nodes;

i. Monitor the transmission status of the local network, receive and summarize the transmission status of the network under its jurisdiction, and report to the information server;

j. Receive the traffic management policy of the nodes within the jurisdiction analyzed by the policy server, and distribute the policy to common nodes;

k. Realize local traffic management according to the received local traffic management policy.

The authentication server in the functional entity implements the function of client user access network authority authentication. The server stores user authentication information of all clients to realize new user registration and password modification; specific information interaction includes: receiving registration information of newly added nodes, receiving all The password modification information of the client, conveys the newly joined node information to the directory server.

The directory server in the functional entity saves all super node information, allocates the list of super nodes for newly added ordinary nodes, communicates the ordinary node information of the requested super node to the corresponding assigned super node, and communicates the list of all super node information to the policy server and information servers, so that policy servers can distribute policies and information servers can summarize network transmission information.

The policy server in the functional entity formulates traffic management and control policies based on load feedback for different policy request points according to the overall network transmission status obtained by the information server, including the load status of each part of the network, and communicates the policy to the requesting policy. The super node, in turn, is communicated by the super node to the ordinary node requesting the policy.

The information server receives the network transmission information conveyed by the super nodes, summarizes and analyzes all the network transmission information in the jurisdiction of each super node, and makes statistics, and transmits the statistical results of the network transmission information to the policy server, so that the policy server can base on the current load of the network State, formulate a traffic management strategy based on load feedback.

Beneficial effect: the advantages of the flow management method are:

1. Efficient and accurate traffic identification

Because this method has corresponding client software installed on all user computers in the network, local traffic identification and management can be realized through this software, which can be accurately identified according to the local process, and traffic identification can be carried out on the network egress device through the characteristics of the data flowing through it. The feasibility is high, and the identification is accurate and efficient.

2. Reduce the load and cost of traffic management equipment

The method includes two parts: client software and special server. As for the client software, it only needs to be installed on the computer of the network user as the identity of the network access pass, without any special configuration; It is enough to configure one or more servers with the same function. Since the super node weakens the function of the server and reduces the server load, and does not need to bear all the network traffic like traditional flow pipe equipment, it is not too high for special server configuration. Requirements, thus greatly reducing the cost of traffic management as a whole.

3. Able to realize network traffic management

The traffic identification and management of this method is not realized by deploying traffic management equipment at the network exit or branch, but by the client software on each user in the network. All users in the network can access the local network and transmit data. Carry out effective traffic management and control, so that the traffic management and control between users in the network can be effectively realized.

Description of drawings

FIG. 1 is a topology diagram of a P2P-based distributed traffic management method. The figure shows the entire network topology of the method, including the main functional entities of the method.

FIG. 2 is a functional hierarchical structure diagram of a P2P-based distributed traffic management method. The figure shows a three-layer functional architecture according to the entity and function division of the method.

Fig. 3 is a flow chart of the client software of the P2P-based distributed flow management method. The figure shows the entire execution flow of the client software in the method.

Detailed ways

The present invention proposes a method for implementing distributed flow management using the P2P principle, which introduces the overall structure and implementation principle of the method, proposes each functional entity included in the P2P-based distributed flow management method, and introduces the structure of each functional entity The use of interaction mechanisms, especially feedback mechanisms, in the method. The method adopts the third-generation P2P architecture, that is, hybrid P2P implementation, which is similar to the third-generation P2P systems such as Skype, BT, etc. This method includes three types of functional entities: ordinary nodes, super nodes and special servers. Different entities have different configurations. In the method In the method, ordinary nodes and super nodes are both ordinary network users. As a network access pass, the corresponding functional software is installed on the computers of ordinary network users. The software is determined according to the configuration of the user's computer. It is also the function of a super node; the special server in the method is one or more servers with different functions configured in the network, including a traffic management policy server, a network information server, and the like. In addition, the method adopts a feedback mechanism, and traffic management strategies are different according to different network load conditions.

The following introduces the overall architecture and working principle of the P2P-based distributed traffic management method, each functional entity and the interaction of the entities.

Firstly, the overall architecture and working principle of the P2P-based distributed traffic management method are introduced.

The method adopts the third-generation hybrid P2P network architecture, which has a functional entity similar to Skype, and has a three-layer functional structure. From bottom to top, there are ordinary node layer, super node layer and server layer, which correspond to three types of functional entities.

All users in the network must install software with local traffic identification and management functions with the identity of the network ID card to access the network. That is, if each user in the network needs to access the Internet, he must first install the software, similar to Dr.COM broadband login Client, and according to the performance of the user's computer, the software automatically determines whether the user belongs to a normal node or a super node. For ordinary nodes and super nodes, the software has basic functions such as network access, traffic identification, and traffic management. information and submit it to the information server and other special functions.

In order to provide effective management for ordinary nodes and super nodes, the method contains a number of special servers. These servers do not manage network traffic like traditional traffic management equipment, nor are they the outlet of network data. They are only used to manage common Nodes and super nodes, and provide necessary information for super nodes or other servers, such as directory server provides super node list information for other special servers, policy server provides traffic management policies for super nodes, information server summarizes network load provided by super nodes, etc. information and submit it to the policy server.

The functional entities and entity interactions contained in the method are introduced in detail below.

1. Ordinary nodes:

Ordinary nodes are network users who have installed client software but have relatively poor computer performance. It has the most basic functions of client software. The specific working process includes:

1) User authentication for local access to the network. Only by using the client for user information authentication can the user have access to the network. The newly added client connects to the authentication server for user authentication, and then the authentication server communicates the user information to the common node The super node to which it belongs is managed by the super node;

2) After obtaining the network access user authentication, the directory server obtains the locally accessible supernode list information, and establishes a connection with the supernode;

3) Obtain the network access and data transmission status of all local active processes, and submit the local network transmission status to the super node;

4) Effectively manage and control the network access and data transmission of all local processes according to the traffic management strategy conveyed by the super node, and realize traffic shaping.

2. Super node:

Super nodes are network users who have installed client software and have relatively good computer performance. In addition to the basic functions of ordinary node clients, it also has the functions of managing ordinary nodes, communicating network transmission information in sub-domains, and communicating traffic management policies, so as to finally achieve the purpose of slowing down the load of special servers in the center.

Compared with ordinary nodes, super nodes have the following characteristics;

1) User authentication for accessing the network is directly connected to the authentication server, and the newly added super node information is communicated to the directory server by the authentication server;

2) When the directory server communicates the super node information to the newly added common child node, it transmits the basic information of the child node to the super node;

3) Summarize the transmission information of each common node, and convey the network transmission information within the jurisdiction to the information server;

4) The traffic management policy is obtained by the policy server, and communicated to the common nodes in the domain requesting the policy.

3. Special server:

Special servers are special entities within the network, similar to login servers and event servers in Skype. According to actual needs, the special server designed by this method includes a total of four special servers: authentication server, directory server, policy server, and information server. The specific functions are as follows:

1) Authentication server:

Realize the function of client user access network authority authentication, store all client user authentication information on the server, and have functions such as new user registration and password modification. The specific information interaction includes: receiving the registration information of the newly added node, receiving the password modification information of all clients, and conveying the information of the newly added node to the directory server, etc.

2) Directory server:

The directory server saves all super node information, can assign the list of super nodes to newly added ordinary nodes, can communicate the ordinary node information of the requested super node to the corresponding assigned super node, and communicate the list of all super node information to the policy server and The information server facilitates the policy server to distribute policies and the information server to summarize network transmission information.

3) Policy server:

According to the overall transmission status of the network obtained by the information server, including the load status of each part of the network, the policy server formulates traffic management and control policies based on load feedback for different policy request points, and communicates the policy to the super node requesting the policy. In turn, the super node communicates to the ordinary node requesting the policy.

4) Information server:

The main function of the information server is to receive the network transmission information conveyed by the super nodes, summarize and analyze all the network transmission information within the jurisdiction of each super node, and make statistics, and pass the statistical results of the network transmission information to the policy server, so that the policy server can follow the According to the current load status of the network, a traffic management strategy based on load feedback can be formulated, so that each client can assign different permissions and traffic shaping to the local process access network according to the current network transmission status, and finally achieve full and reasonable use of network resources the goal of.

The specific application of the P2P-based distributed traffic management method is illustrated below through a simple example.

The traffic management system designed according to the above method can be deployed in a local area network such as a campus network, an enterprise network, or a certain telecommunication network. With the identity of the network access pass, users are forced to install client software on their computers, so that each network user can independently realize the function of network traffic management. Client software needs to be able to be installed across platforms to satisfy network users with different platforms.

The working process of the system is described as follows:

1) A new user installs the client software, generates a special server list, and judges the configuration status of the local computer when running for the first time after installation, so as to determine whether it is a normal node or a super node;

2) Communicate local basic information and new user registration information to the authentication server. If it is a common node, the directory server will obtain the list of super nodes to which it belongs, and the directory server will communicate the general node information to the super node to which it belongs, and finally establish a relationship with the super node. Node connection; if it is a super node, the authentication server will communicate the super node information to the directory server to update the super node list;

3) When ordinary nodes have network access and data transmission behaviors, they report the local network status to super nodes, and obtain traffic management policies through super nodes;

4) The super node communicates the network transmission information of the ordinary nodes under its jurisdiction to the information server, and the information server feeds back to the policy server, and the policy server communicates the traffic management policy to the super node, and the super node communicates it to the ordinary node requesting the policy;

5) The authentication server updates the authentication information of all users on the network in a timely manner, and constantly exchanges information with the directory server to update the super node list information stored in the directory server;

6) The information server receives the network transmission information conveyed by each super node in a timely manner, and feeds back the network transmission status to the policy server, so as to formulate a traffic management strategy with a feedback mechanism, ensure the rationality of policy formulation, and fully and rationally use network resources.

After the system is deployed, users in the network can accurately identify local network services, automatically adopt different strategies according to the current network load status, and realize network traffic management and control, which not only improves the satisfaction of users in the network, but also makes full use of network resources , prevent resource waste, and reduce network operation and maintenance costs.

Claims (7)

1. the distributed stream quantity management method based on peer-to-peer network is characterized in that this method adopts third generation peer-to-peer network framework, promptly adopts the peer-to-peer network principle to realize distributed network flow identification and management at each network user end; This methodological function entity comprises ordinary node, super node and certificate server, LIST SERVER, strategic server and information server; In addition, this method adopts feedback mechanism, formulates flow management strategy according to the offered load situation, and concrete flow managing method is:

1) user installation client software in the net generates the special server tabulation, judges the local computer configuration state when moving first after the installation, thereby determines ordinary node or super node;

2) client software operation, local essential information and new user's registration information are conveyed to certificate server, if ordinary node, obtain affiliated super node tabulation by LIST SERVER, and by LIST SERVER with the super node of this ordinary node information under conveying to, final foundation and being connected of super node; If super node conveys to LIST SERVER by certificate server with this super node information and upgrades the super node tabulation;

When 3) ordinary node has access to netwoks and transfer of data behavior, give super node, and obtain flow management strategy by super node with the local network status report;

4) super node conveys to information server with the Network Transmission information of linchpin ordinary node, feed back to strategic server by information server, strategic server conveys to super node with flow management strategy and is conveyed to the ordinary node of request strategy by super node;

5) certificate server all users' of network the authentication information that upgrades in time, and constantly carry out information interaction, the super node list information that the renewal LIST SERVER is stored with LIST SERVER;

6) information server in time receives the Network Transmission information that each super node is passed on, and give strategic server with the Network Transmission condition feedback, thereby formulate flow management strategy, guarantee the reasonability of policy development, sufficiently and reasonably utilize Internet resources with feedback mechanism.

2. the distributed stream quantity management method based on peer-to-peer network according to claim 1 is characterized in that the overall architecture that realizes this method is:

1) adopts third generation mixed P 2 P network framework, have the functional entity that is similar to Skype;

2) having three layer function structures, is ordinary node layer, super node layer and server layer from bottom to up successively, corresponding successively three class functional entitys;

3) carrying out access to netwoks by user in all nets must install the software with local flow identification and management function and realize with the identity of networking identity card;

4) adopt feedback mechanism, formulate flow management strategy according to the network in general transmission state.

3. the distributed stream quantity management method based on peer-to-peer network according to claim 1 is characterized in that the implementation of the client software that ordinary node and super node are installed is:

1) operation of client software at first judges whether to move first, if then check the local computer configuration state, judges whether the user can become super node,

If a. meet super node, then to certificate server registration super node, and upgrade the super node tabulation by certificate server notice LIST SERVER, non-to move implementation first identical with super node afterwards;

If b. do not meet super node, then to certificate server registration ordinary node, and connect LIST SERVER and obtain nearest super node tabulation, non-to move implementation first identical with ordinary node afterwards;

2) the non-operation first of client software, if ordinary node, then:

C. surveying super node tabulates and connects with certain super node;

D. the authentification of user by super node, and obtain network access authority;

E. monitor local current network status transmission and report to super node;

F. the policy information that receives the super node transmission is realized local traffic management and control;

3) the non-operation first of client software, if super node, then:

G. the authentification of user by certificate server, and obtain network access authority;

H. intercept and receive the connection of ordinary node;

I. monitor the local network status transmission, receive and gather institute's scope Network Transmission situation, report to information server;

J. receive the flow management strategy of institute's scope interior nodes that the strategic server analysis draws, and give ordinary node strategy distribution;

K. realize local traffic management according to the local flow management strategy that receives.

4, the distributed stream quantity management method based on peer-to-peer network according to claim 1, it is characterized in that certificate server is realized the function that the client user accesses network legal power authenticates in the functional entity, the user authentication information of all clients of storage on the server is realized new user's registration, password modification; Specifying information comprises alternately: receive the log-on message of newly added node, the password modification information that receives all clients, reception and registration newly added node information to LIST SERVER.

5, the distributed stream quantity management method based on peer-to-peer network according to claim 1, it is characterized in that LIST SERVER is preserved all super node information in the functional entity, for the new ordinary node that adds distributes affiliated super node tabulation, the ordinary node information of request super node is conveyed to the corresponding super node that distributes, and all super node information lists are conveyed to strategic server and information server, thereby be convenient to the strategic server allocation strategy and information server gathers Network Transmission information.

6, the distributed stream quantity management method based on peer-to-peer network according to claim 1, it is characterized in that strategic server in the functional entity is according to network in general status transmission that information server obtained, the load state that comprises network each several part inside, formulation is at the traffic management and the control strategy based on load feedback of Different Strategies request point, and strategy conveyed to the super node of request strategy, and then convey to the ordinary node of request strategy by super node.

7, the distributed stream quantity management method based on peer-to-peer network according to claim 1, it is characterized in that information server receives the Network Transmission information that super node is passed on, all-network transmission information in each super node institute scope is carried out Macro or mass analysis, and add up, the statistics of Network Transmission information is passed to strategic server, be convenient to the strategic server load state current, formulate flow management strategy based on load feedback according to network.