CN101883052A - Method and system for realizing traffic optimization in peer-to-peer network - Google Patents

Abstract

The invention discloses a method and a system for realizing traffic optimization in a peer-to-peer network. The method comprises the following steps of: performing linear weighing on bottom characteristic parameters affecting communication overhead to obtain a total overhead value of links among different nodes; and determining a network priority sequence containing node information according to the obtained total overhead value, wherein the node information indicates nodes of a client for communication interaction. Compared with the prior art, the method and the system for realizing traffic optimization in the peer-to-peer network do not need to restrain P2P traffic, have no problems or defects of complex deployment equipment and the like, localize the P2P traffic to reduce cross-domain traffic, can process different P2Ps, coordinate diversity and particularity of P2P services, and optimize user experience of the P2P.

Description

A kind of method and system of realizing flow optimization in the peer-to-peer network

Technical field

The present invention relates to the communications field, be specifically related to a kind of method and system of realizing flow optimization in the peer-to-peer network.

Background technology

P2P (peer to peer) i.e. equity calculates or peer-to-peer network, can be defined as the mode by direct exchange simply, shares computer resource and service.In the P2P network environment, thousands of the computers that are connected with each other all are in reciprocity status, and whole network does not rely on special-purpose centralized servers in general.Each computer can serve as the requestor of network service in the network, also can make response to the request of other computers, and resource and service are provided.Based on above characteristics, P2P has become the main means that mass datas such as multimedia are provided for the user on the broadband networks.

But the network that is extensive use of to operator of P2P application has in recent years brought white elephant.On the long-distance backbone network of Chinese operator, the P2P flow accounts for over half, has had a strong impact on the communication effect with other traditional mainstream content of carrying out of the Internet regular traffic.This is because due to the chronic illness of P2P algorithm.At present P2P flow optimization, some solutions have been proposed at present.As: by message flow is carried out pattern matching, restriction strategy is combined with access control list ACL, the P2P message is carried out various current limlitings operate and retrain the P2P flow.But this method is a cost with the experience of sacrificing P2P user, can't satisfy the demand of P2P development.And for example: download detection and control subsystem, P2P file cache and distribution subsystem by disposing P2P, beam split detects identification P2P flow, and file cache is optimized the P2P flow to this locality.But this method depends on the passive detection and the buffer memory of P2P flow, needs the equipment of deployment too much, is difficult to realize lightweight and standardized optimum management, optimizes nor can should be used for distinguishing to different P2P.

Summary of the invention

In view of this, main purpose of the present invention is to provide a kind of method and system of realizing flow optimization in the peer-to-peer network, reduces the cross-domain flow rate of P2P.

For achieving the above object, technical scheme of the present invention is achieved in that

A kind of method that realizes flow optimization in the peer-to-peer network, this method comprises:

Carry out linear weighted function at the low-level image feature parameter that influences communication overhead, obtain the overhead value of link between different nodes;

Determine to comprise the network priority sequence of nodal information according to the overhead value that obtains, this nodal information indicates client can communicate mutual node.

Described low-level image feature parameter comprise following one of at least:

Bandwidth, time delay, packet loss, hop count.

When described low-level image feature parameter comprised bandwidth, time delay, packet loss and hop count, the method for described linear weighted function was:

$\mathrm{Cost}=[\frac{K_1}{\mathrm{BandWidth}\left(\mathrm{kbps}\right)}\×{\mathrm{\μ}}_1+\left(\mathrm{\ΣDelay}\left(\mathrm{\μs}\right)\right)\×K_2\×{\mathrm{\μ}}_2+\mathrm{Lost}\×K_3\×{\mathrm{\μ}}_3]+\mathrm{Dis}\tan \mathrm{ce}\×K_0\×{\mathrm{\μ}}_0;$

Wherein, K _i* μ _iBe respectively the weight coefficient of bandwidth, time delay, packet loss, hop count, μ ₀, μ ₁, μ ₂, μ ₃Sum is 100%, K _iBe μ _iNormalized parameter.

The method of determining described network priority sequence is:

Content ascending order in the described overhead value sequence is arranged, obtains the network priority sequence of following expression:

$\left[\begin{array}{c}\mathrm{RNetID}11\\ \mathrm{RNetID}12\\ \mathrm{RNetID}13\\ \·\·\·\·\·\·\·\end{array}\right]=\left[\begin{array}{cccc}{n}_1\mathrm{ID}{r}_1& n_1\mathrm{ID}{r}_2& n_1\mathrm{ID}{r}_3& \·\·\·\\ n_2\mathrm{ID}{r}_1& n_2\mathrm{ID}{r}_2& n_2\mathrm{ID}{r}_3& \·\·\·\\ n_3\mathrm{ID}{r}_1& n_3\mathrm{ID}{r}_2& n_3\mathrm{ID}{r}_3& \·\·\·\\ \·\·\·& \·\·\·\·& \·\·\·& \·\·\·\end{array}\right];$

Wherein, n _aIDr _bRepresent that certain node is in the network that ID is a; When this node sends when request, be the network ID number of b for the priority of its selection.

This method further comprises:

Nodal information in the described network priority sequence of client feedback that proposes query requests, client is initiated content requests to the pairing node of the nodal information of receiving.

A kind of system that realizes flow optimization in the peer-to-peer network, this system comprise expense decision package, network priority sequence decision package; Wherein,

Described expense decision package is used for carrying out linear weighted function at the low-level image feature parameter that influences communication overhead, obtains the overhead value of link between different nodes;

Described network priority sequence decision package, the overhead value that is used for obtaining according to the expense decision package determines to comprise the network priority sequence of nodal information; Described nodal information indicates client can communicate mutual node.

Described low-level image feature parameter comprise following one of at least: bandwidth, time delay, packet loss, hop count;

Described expense decision package, network priority sequence decision package are arranged in the P2P re-positioning device.

When described low-level image feature parameter comprised bandwidth, time delay, packet loss and hop count, described expense decision package was used to carry out following calculating when carrying out linear weighted function:

$\mathrm{Cost}=[\frac{K_1}{\mathrm{BandWidth}\left(\mathrm{kbps}\right)}\×{\mathrm{\μ}}_1+\left(\mathrm{\ΣDelay}\left(\mathrm{\μs}\right)\right)\×K_2\×{\mathrm{\μ}}_2+\mathrm{Lost}\×K_3\×{\mathrm{\μ}}_3]+\mathrm{Dis}\tan \mathrm{ce}\×K_0\×{\mathrm{\μ}}_0;$

Wherein, K _i* μ _iBe respectively the weight coefficient of bandwidth, time delay, packet loss, hop count, μ ₀, μ ₁, μ ₂, μ ₃Sum is 100%, K _iBe μ _iNormalized parameter.

Described network priority sequence decision package is used for the content ascending order of described overhead value sequence is arranged when determining described network priority sequence, obtains the network priority sequence of following expression:

$\left[\begin{array}{c}\mathrm{RNetID}11\\ \mathrm{RNetID}12\\ \mathrm{RNetID}13\\ \·\·\·\·\·\·\·\end{array}\right]=\left[\begin{array}{cccc}{n}_1\mathrm{ID}{r}_1& n_1\mathrm{ID}{r}_2& n_1\mathrm{ID}{r}_3& \·\·\·\\ n_2\mathrm{ID}{r}_1& n_2\mathrm{ID}{r}_2& n_2\mathrm{ID}{r}_3& \·\·\·\\ n_3\mathrm{ID}{r}_1& n_3\mathrm{ID}{r}_2& n_3\mathrm{ID}{r}_3& \·\·\·\\ \·\·\·& \·\·\·\·& \·\·\·& \·\·\·\end{array}\right];$

Wherein, n _aIDr _bRepresent that certain node is in the network that ID is a; When this node sends when request, be the network ID number of b for the priority of its selection.

This system comprises that further nodal information provides the unit, is used for the nodal information to the described network priority sequence of client feedback that proposes query requests;

Described client is used for the pairing node of the nodal information of receiving is initiated content requests.

The present invention realizes the method and system of flow optimization in the peer-to-peer network, compared with prior art, needn't retrain the P2P flow, does not have problem and defectives such as deployment facility complexity, but P2P is flow localized, has reduced cross-domain flow rate; And can should be used as different processing to different P2P, and take into account P2P diversity of operations and particularity, optimized P2P user's experience.

Description of drawings

Fig. 1 is the system diagram of flow optimization in the realization peer-to-peer network of one embodiment of the invention;

Fig. 2 is the abstract schematic diagram of the network topology of one embodiment of the invention;

Fig. 3 is the process chart that utilizes low-level image feature parameter optimization node to select of one embodiment of the invention;

Fig. 4 realizes the general flow chart of flow optimization in the peer-to-peer network for the present invention.

Embodiment

In general, can be by concrete quantification, tolerance network topological information, interconnected according to the professional attribute information guiding node of submitting to of P2P, when obviously reducing operator's burden, farthest optimize the user experience of multiple P2P business.

When concrete the application, the deployment architecture figure of P2P flow optimization in the single domain can be provided as shown in Figure 1, similar deployment form also can be adopted in other reciprocity territory.Referring to Fig. 1, Fig. 1 is the system diagram of flow optimization in the realization peer-to-peer network of one embodiment of the invention.

Wherein, PPR (P2P Redirector, P2P re-positioning device) is as the network entity that is deployed in the autonomous territory of operator, can be by open standard agreement and the P2P client communication in the territory, and for the P2P user in this territory provides service.In case received the peer table request that the P2P client sends, will select the Peer table to return according to network topological information and policy information; DNS (domain name analysis system) provides the domain name mapping service for client, client can be in its inquiry field the actual address of PPR; PPC (P2P Cache, P2P content caching equipment) is deployed in the territory as a super seed, for P2P user in the territory provides agency and caching function.PPC can rationally dispose according to the existing situation of network, and inessential deployment facility.

Because PPR is deployed in the territory, therefore can obtains the topology information of network reality and set up network model.According to the actual conditions of network and the granularity of network management, network can be divided into different territories, may there be multi-level subnet in each territory down.Operator is deposited into its topology information in the database with specific form.With the visual angle of PPR, can change into an oriented connected graph with whole network is abstract, Fig. 2 is the abstract schematic diagram of the present invention's network topology when carrying out network measure.

In realizing peer-to-peer network, during flow optimization, can carry out step as follows substantially:

The first step, the client that operation P2P uses is by the address of PPR in the DNS inquiry field;

In second step, client is initiated registration to PPR, submits to described P2P to use the weight coefficient of the concrete low-level image feature parameter of being paid close attention to;

In the 3rd step, PPR carries out linear weighted function at the low-level image feature parameter that influences communication overhead, obtains the overhead value of link between different nodes; Determine to comprise the network priority sequence of nodal information according to the overhead value that obtains.Described nodal information is information such as node ID, IP, is used in reference to client is shown can communicates mutual node (this node can be called the alternative services node).

In the 4th step, client is to the query requests of PPR submission at the alternative services node, and PPR is to nodal informations such as the node ID in its answer network priority sequence, IP;

In the 5th step, client is initiated content requests to the pairing node of the nodal information of receiving.

Need to prove, expense decision package, network priority sequence decision package, nodal information can be set in functional entitys such as PPR the unit is provided.Wherein, the expense decision package can carry out linear weighted function at the low-level image feature parameter that influences communication overhead, obtains the overhead value of link between different nodes; Network priority sequence decision package can determine to comprise the network priority sequence of nodal information according to the overhead value that the expense decision package obtains; Nodal information provides the unit then can make client to initiate content requests to the pairing node of the nodal information of receiving to the nodal information in the described network priority sequence of client feedback that proposes query requests.

During concrete the application,,, need to create following three tables to optimize the selection of node according to parameter in order to measure the parameter between the different nodes:

First table is Net (network) table, in order to describe the attribute in territory.Basic list item is＜NetID ParentID, Type, AddrStart, AddrEnd, Mask, Capacity 〉.Network hierarchical relationship each other can be by＜NetID, ParentID〉two parameters decide; Type is used for explaining the type and the access way of network; AddrStart and AddrEnd are used for representing to have in this network the IP scope of node; Mask represents subnet mask; Capacity is used for representing the interstitial content that network is interior.

Second table is the Peer table, in order to describe the base attribute of territory interior nodes.Basic list item is＜PeerID IP, PeerUplink, PeerDownlink, Option 〉.PeerID and IP represent the logical identifier and the network layer sign of this node respectively; PeerUplink and PeerDownlink represent uploading of this node and download bandwidth respectively; Option is used for doing scaling option, can represent situations such as the load of computer and computational resource.

The 3rd table is the NetEage table, is used to represent the parameter information of network link between each network.Basic list item is＜NetA NetB, BandWidth, Delay, Lost, Distance, Option 〉.NetA and NetB are network ID number, represent which two network this link connects; BandWidth represents the outlet bandwidth of network; Delay represents the time delay of link; Lost represents the packet loss and the reliability of link; Distance represents the router number that this link is crossed over; Options is a scaling option.

It needs to be noted: Fig. 2 is a directed graph.In each network＜NetA, NetB〉and＜NetB, NetA〉represent the different directions of consolidated network link respectively, because parameters such as its outlet bandwidth of the different directions of same link and time delay may be different.

Consider that operator and P2P application need carry out the optimized choice of node at the different emphasis of each parameter, can be with the low-level image feature parameter in the NetEage table as independent variable, take all factors into consideration the weight coefficient of each low-level image feature parameter, calculate the overhead value Cost of link between the different nodes.Carrying out just can selecting the network segment and node when node is selected according to the Cost of different P2P business.As, carry out linear weighted function at the low-level image feature parameter that influences communication overhead, obtain the overhead value of link between different nodes, formula is as follows:

$\mathrm{Cost}=[\frac{K_1}{\mathrm{BandWidth}\left(\mathrm{kbps}\right)}\×{\mathrm{\μ}}_1+\left(\mathrm{\ΣDelay}\left(\mathrm{\μs}\right)\right)\×K_2\×{\mathrm{\μ}}_2+\mathrm{Lost}\×K_3\×{\mathrm{\μ}}_3]+\mathrm{Dis}\tan \mathrm{ce}\×K_0\×{\mathrm{\μ}}_0;$

Wherein, K _i* μ _iBe respectively the weight coefficient of bandwidth, time delay, packet loss, hop count, μ ₀, μ ₁, μ ₂, μ ₃Sum is 100%, K _iBe μ _iNormalized parameter.

In above each weight coefficient, μ ₀Main operatable object merchant is in order to regulate the size of Distance weight.The hop count of Distance for being crossed on the path of certain bar Route Selection.Operator wishes that the Distance value of selected node is as far as possible little, thereby reduces cross-domain flow rate, and this just need promote coefficient μ when the Cost computing formula is set ₀Size.μ ₀Be not fixed value, have different values, but concrete setting still needs to do between P2P manufacturer and the operator and gos deep into concrete negotiation more but use at different P2P.

μ ₁, μ ₂And μ ₃Adjusting mainly use towards P2P, be used for representing the size of bandwidth, time delay and coefficient of reliability (packet loss) respectively, time delay wherein need be taken all factors into consideration the time delay of link itself and the time delay of the network equipment.Different P2P service needed is selected different＜μ ₁, μ ₂, μ ₃Vector, for the business of delay sensitives such as similar Skype voice, need to strengthen μ ₂Coefficient; For downloading services such as Bittorrent, just need the selection coefficient μ of increased bandwidth ₁For streaming media service such as videos, then need to take all factors into consideration the influence of factors such as bandwidth and time delay.

For the first time that each is concrete P2P uses when PPR registers, and needs to submit to the service attribute of P2P application, comprises＜μ ₁, μ ₂, μ ₃Vector value.When carrying out this professional client once more to the PPR requesting node, based on different business, PPR can calculate the Cost of node to be selected network of living in according to the vector value of registration before, selects thereby optimize node.Afterwards, after setting the required weight coefficient of calculating overhead value, PPR can adopt traditional dijkstra's algorithm to generate the Cost sequence of each network to other all-networks, obtains the result of calculation that presents with following matrix-style after the content ascending order in this sequence is arranged:

$\left[\begin{array}{c}P2\mathrm{PType}1\\ P2\mathrm{PType}2\\ P2\mathrm{PType}3\\ \·\·\·\·\·\·\·\end{array}\right]=\left[\begin{array}{cccc}\mathrm{RNetID}11& \mathrm{RNetID}12& \mathrm{RNetID}13& \·\·\·\\ \mathrm{RNetID}21& \mathrm{RNetID}22& \mathrm{RNetID}23& \·\·\·\\ \mathrm{RNetID}31& \mathrm{RNetID}32& \mathrm{RNetID}33& \·\·\·\\ \·\·\·& \·\·\·\·& \·\·\·& \·\·\·\end{array}\right];$

Wherein, NetIDi represents to be the network of i ID number; RNetIDmn represents that the P2P type is m, and being at ID is in the network of n.

PPR can be expressed as according to the network priority sequence that Cost calculates:

$\left[\begin{array}{c}\mathrm{RNetID}11\\ \mathrm{RNetID}12\\ \mathrm{RNetID}13\\ \·\·\·\·\·\·\·\end{array}\right]=\left[\begin{array}{cccc}{n}_1\mathrm{ID}{r}_1& n_1\mathrm{ID}{r}_2& n_1\mathrm{ID}{r}_3& \·\·\·\\ n_2\mathrm{ID}{r}_1& n_2\mathrm{ID}{r}_2& n_2\mathrm{ID}{r}_3& \·\·\·\\ n_3\mathrm{ID}{r}_1& n_3\mathrm{ID}{r}_2& n_3\mathrm{ID}{r}_3& \·\·\·\\ \·\·\·& \·\·\·\·& \·\·\·& \·\·\·\end{array}\right];$

Wherein, n _aIDr _bRepresent that certain node is in the network that ID is a; When this node sends when request, PPR is the network ID number of b for the priority of its selection.

In the foregoing description, considered the measurement to link between net, promptly the parameter in the Cost computing formula comes from the NetEage table.The Peer table is the parametric description to node self, does not embody temporarily in Cost calculates.The Net table is the description to existing network in the network, and the meaning of existence is arranged, but does not temporarily calculate meaning.

Referring to Fig. 3, Fig. 3 is the process chart that utilizes low-level image feature parameter optimization node to select of one embodiment of the invention, and this flow process may further comprise the steps:

Step 301: when client is initiated request, judge whether to inquire the pairing P2P type of the request of receiving,, enter step 302 if can inquire; Otherwise, enter step 310.

Step 302: judge whether to inquire the pairing fileinfo of the request of receiving,, enter step 303 if can inquire; Otherwise, enter step 320.

Step 303:, determine the overhead value of link between the different nodes according to type of service and the network that is requested the node place of P2P.

Step 304: return the peer table, process ends.

Step 310: return the peer table for empty, process ends.

Step 320: judge whether the request that client is initiated is request for the first time, if enter step 330; Otherwise, return step 310.

Step 330: ask pairing file content to add the file table to this, and upgrade corresponding Net table and Peer table.

As seen from Figure 3, when client when PPR initiates request, PPR can judge whether to exist the pairing node listing of described request according to the type of client P2P application and the content resource of collecting, and when having node listing, utilize the Cost between each territory that PPR calculates preferentially to select to have the territory interior nodes of less Cost and return to client.

In deployment examples of the present invention, not only considered benefits of operators, also taken into account multiple P2P user's experience.Be applied as example with the Bittorrent application of download class and the PPCDN of Streaming Media class, Bittorrent uses mainly concern hop count and these two parameters of bandwidth, obtains best downloading experience in the time of in the hope of the minimizing cross-domain flow rate; Comparatively speaking, the bandwidth of 800k/s can satisfy the transmission requirement of most of Streaming Media channels substantially, and PPCDN then more pays close attention to parameters such as packet loss and time delay so, so that better meet client's demand.Be applied in the deployment examples of the present invention with this two class and carry out experiment, all be subjected to tangible optimization effect.

In conjunction with above description as can be known, the present invention realizes that the operation thinking of flow optimization in the peer-to-peer network can represent as shown in Figure 4.Referring to Fig. 4, Fig. 4 realizes the general flow chart of flow optimization in the peer-to-peer network for the present invention, and this flow process may further comprise the steps:

Step 410: carry out linear weighted function at the low-level image feature parameter that influences communication overhead, obtain the overhead value of link between different nodes.

Step 420: determine to comprise the network priority sequence of nodal information according to the overhead value that obtains, this nodal information indicates client can communicate mutual node.

In sum as seen, no matter be method or system, adopt the technology of flow optimization in the realization peer-to-peer network of the present invention, compared with prior art, needn't retrain the P2P flow, not have problem and defectives such as deployment facility complexity, but P2P is flow localized, reduced cross-domain flow rate; And can should be used as different processing to different P2P, and take into account P2P diversity of operations and particularity, optimized P2P user's experience.

The above is preferred embodiment of the present invention only, is not to be used to limit protection scope of the present invention, all any modifications of being done within the spirit and principles in the present invention, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.

Claims (10)

1. a method that realizes flow optimization in the peer-to-peer network is characterized in that, this method comprises:

Carry out linear weighted function at the low-level image feature parameter that influences communication overhead, obtain the overhead value of link between different nodes;

Determine to comprise the network priority sequence of nodal information according to the overhead value that obtains, this nodal information indicates client can communicate mutual node.

2. method according to claim 1 is characterized in that, described low-level image feature parameter comprise following one of at least:

Bandwidth, time delay, packet loss, hop count.

3. method according to claim 2 is characterized in that, when described low-level image feature parameter comprised bandwidth, time delay, packet loss and hop count, the method for described linear weighted function was:

$\mathrm{Cost}=[\frac{K_1}{\mathrm{BandWidth}\left(\mathrm{kbps}\right)}\×{\mathrm{\μ}}_1+\left(\mathrm{\ΣDelay}\left(\mathrm{\μs}\right)\right)\×K_2\×{\mathrm{\μ}}_2+\mathrm{Lost}\×K_3\×{\mathrm{\μ}}_3]+\mathrm{Dis}\tan \mathrm{ce}\×K_0\×{\mathrm{\μ}}_0;$

Wherein, K _i* μ _iBe respectively the weight coefficient of bandwidth, time delay, packet loss, hop count, μ ₀, μ ₁, μ ₂, μ ₃Sum is 100%, K _iBe μ _iNormalized parameter.

4. method according to claim 3 is characterized in that, determines that the method for described network priority sequence is:

Content ascending order in the described overhead value sequence is arranged, obtains the network priority sequence of following expression:

$\left[\begin{array}{c}\mathrm{RNetID}11\\ \mathrm{RNetID}12\\ \mathrm{RNetID}13\\ \·\·\·\·\·\·\·\end{array}\right]=\left[\begin{array}{cccc}{n}_1\mathrm{ID}{r}_1& n_1\mathrm{ID}{r}_2& n_1\mathrm{ID}{r}_3& \·\·\·\\ n_2\mathrm{ID}{r}_1& n_2\mathrm{ID}{r}_2& n_2\mathrm{ID}{r}_3& \·\·\·\\ n_3\mathrm{ID}{r}_1& n_3\mathrm{ID}{r}_2& n_3\mathrm{ID}{r}_3& \·\·\·\\ \·\·\·& \·\·\·\·& \·\·\·& \·\·\·\end{array}\right];$

Wherein, n _aIDr _bRepresent that certain node is in the network that ID is a; When this node sends when request, be the network ID number of b for the priority of its selection.

5. according to each described method of claim 1 to 4, it is characterized in that this method further comprises:

Nodal information in the described network priority sequence of client feedback that proposes query requests, client is initiated content requests to the pairing node of the nodal information of receiving.

6. a system that realizes flow optimization in the peer-to-peer network is characterized in that, this system comprises expense decision package, network priority sequence decision package; Wherein,

Described expense decision package is used for carrying out linear weighted function at the low-level image feature parameter that influences communication overhead, obtains the overhead value of link between different nodes;

Described network priority sequence decision package, the overhead value that is used for obtaining according to the expense decision package determines to comprise the network priority sequence of nodal information; Described nodal information indicates client can communicate mutual node.

7. system according to claim 6 is characterized in that,

Described low-level image feature parameter comprise following one of at least: bandwidth, time delay, packet loss, hop count;

Described expense decision package, network priority sequence decision package are arranged in the P2P re-positioning device.

8. system according to claim 7 is characterized in that, when described low-level image feature parameter comprised bandwidth, time delay, packet loss and hop count, described expense decision package was used to carry out following calculating when carrying out linear weighted function:

$\mathrm{Cost}=[\frac{K_1}{\mathrm{BandWidth}\left(\mathrm{kbps}\right)}\×{\mathrm{\μ}}_1+\left(\mathrm{\ΣDelay}\left(\mathrm{\μs}\right)\right)\×K_2\×{\mathrm{\μ}}_2+\mathrm{Lost}\×K_3\×{\mathrm{\μ}}_3]+\mathrm{Dis}\tan \mathrm{ce}\×K_0\×{\mathrm{\μ}}_0;$

Wherein, K _i* μ _iBe respectively the weight coefficient of bandwidth, time delay, packet loss, hop count, μ ₀, μ ₁, μ ₂, μ ₃Sum is 100%, K _iBe μ _iNormalized parameter.

9. system according to claim 8 is characterized in that, described network priority sequence decision package is used for the content ascending order of described overhead value sequence is arranged when determining described network priority sequence, obtains the network priority sequence of following expression:

$\left[\begin{array}{c}\mathrm{RNetID}11\\ \mathrm{RNetID}12\\ \mathrm{RNetID}13\\ \·\·\·\·\·\·\·\end{array}\right]=\left[\begin{array}{cccc}{n}_1\mathrm{ID}{r}_1& n_1\mathrm{ID}{r}_2& n_1\mathrm{ID}{r}_3& \·\·\·\\ n_2\mathrm{ID}{r}_1& n_2\mathrm{ID}{r}_2& n_2\mathrm{ID}{r}_3& \·\·\·\\ n_3\mathrm{ID}{r}_1& n_3\mathrm{ID}{r}_2& n_3\mathrm{ID}{r}_3& \·\·\·\\ \·\·\·& \·\·\·\·& \·\·\·& \·\·\·\end{array}\right];$

Wherein, n _aIDr _bRepresent that certain node is in the network that ID is a; When this node sends when request, be the network ID number of b for the priority of its selection.

10. according to each described system of claim 6 to 9, it is characterized in that this system comprises that further nodal information provides the unit, be used for nodal information to the described network priority sequence of client feedback that proposes query requests;

Described client is used for the pairing node of the nodal information of receiving is initiated content requests.

Images