CN103780693A - Optimum replica server selection method and system with user view collaboration awareness used - Google Patents

Abstract

The invention provides an optimum replica server selection method and system with user view collaboration awareness used. The method includes the steps that a user view is established for each user node, and all replica servers which the user nodes in the user views can be connected with are determined; user perspective definition of all the replica servers corresponding to the user nodes in the user views are determined; priorities of all the replica servers corresponding to the user nodes are determined according to the user perspective definition; the information of at least one other user node is obtained through collaboration awareness, and the priorities of the replica servers of the user nodes are adjusted dynamically according to the information of other user nodes. The optimum replica server selection method and system with user view collaboration awareness used can improve the effect on optimum replica server selection of the user nodes and prevent network congestion and server overload from occurring.

Description

A kind of best copy server selection method and system of utilizing User cooperation perception

Technical field

The present invention relates to networking technology area, refer to especially a kind of best copy server selection method and system of utilizing User cooperation perception.

Background technology

Content distributing network (CDN) is near on the Edge Server (also referred to as replica server) close to from user side by the contents distribution on original server, when having need to access these contents time, user needn't read from original server, but directly obtain from replica server, alleviate like this original server owing to accessing the risk of the periods of network disruption causing greatly.

Content distributing network can reduce the load of original server, but owing to existing a large amount of replica servers in network, the best contiguous replica server of How to choose becomes the key issue that reduces offered load and server load.

Network configuration that existing CDN-P2P merges content distributing network as shown in Figure 1, comprises central server layer, replica server layer, user node P2P Autonomous Domain layer.Wherein central server layer comprises original server; Replica server layer comprises a large amount of replica servers; User node P2P Autonomous Domain layer comprises a large number of users node, and in the time that user node reaches some, this network is carried out P2P contents distribution.

Merge in content distributing network at CDN-P2P, although the user in P2P Autonomous Domain can share mutually content resource, but adding of node often has randomness with exiting, and can not provide stable service for user, and therefore replica server remains main ISP.For improving the speed of response user access request, reduce the congested of network channel, select best replica server to become the emphasis of research.Merge in content distributing network at existing CDN-P2P, user node is to select replica server nearby according to the physical location of oneself.And in actual conditions, inhomogeneous due to region and population distribution, the number of users difference in each region, therefore just selects replica server as best copy server according to geographical position, can cause replica server corresponding to region that user node number is many to occur overload phenomenon.And if user node is assigned to geographical position replica server far away as best copy server, can affect data transmission quality.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of method and system that can better select best copy server, and distributing user node that can be more balanced is also guaranteed the data transmission quality of user node.

In order to address the above problem, the embodiment of the present invention has proposed a kind of best copy server selection method that utilizes User cooperation perception, comprising:

Step 1, set up User for each user node, and user node in definite User all replica servers that can connect;

Step 2, determine the user perspective definition of each replica server that each user node in User is corresponding;

Step 3, determine the priority of each replica server that each user node is corresponding according to user perspective definition;

Step 4, obtain the information of at least one other user node by cooperation perception, and dynamically adjust the priority of the replica server of described user node according to the information of described other user nodes.

As technique scheme preferably, described step 2 is specially:

Determine the user perspective definition of each replica server to user node by following formula:

$F_{n_i^c}=\underset{j\∈N\backslash \{i,c\}}{\mathrm{\Σ}}\frac{d(n_i,n_c)d(n_j,n_c)}{d{(n_i,n_j)}^2}$

Wherein,

N _i, n _jrepresent respectively different replica servers;

N _cfor user node;

D (n _i, n _c) be the network delay between user node nc and replica server ni; D (n _j, n _c) be user node n _cwith replica server n _jbetween network delay; D (n _i, n _j) be two internodal network delays of replica server; N is the replica server sum in User, N={n _i| i ∈ { 0 ... k}}.

As technique scheme preferably, described step 3 is specially: determine the priority of each replica server that each user node is corresponding according to user perspective definition F, wherein the priority of higher this replica server of user perspective definition F of replica server is higher.

As technique scheme preferably, described step 4 is specially:

Obtain the priority of replica server of at least one other user node or the best copy server that described other user nodes are selected by cooperation perception, dynamically to adjust the priority of replica server of described user node.

As technique scheme preferably, described step 4 is specially:

Step 41, according to the visual angle definition F of each replica server in User, can calculate user node n _iselect replica server s _jprobability P _ij

$P_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\∈\{\mathrm{1,2}\·\·\·n\}}{\overset{n}{\mathrm{\Σ}}}{F^i}_{S_j}}$

Wherein n _iin the User of individual user node, comprise N replica server s ₁, s ₂, s ₃s _n;

Step 42, user node obtain the priority of the replica server of at least one other user node by cooperation perception, or user node obtains by cooperation perception the best copy server that at least one other user node is selected;

The information that step 43, described user node obtain according to step 42 is dynamic, obtains the user node n after adjustment _iselect replica server s _jprobability p _ir', as this replica server s _jpriority.

Meanwhile, the embodiment of the present invention has also proposed a kind of best copy server selective system of utilizing User cooperation perception, comprising:

User module, is used to each user node to set up User, and user node in definite User all replica servers that can connect;

User perspective definition module, for the user perspective definition of each replica server corresponding to each user node of definite User;

Priority block, for determining the priority of each replica server that each user node is corresponding according to user perspective definition;

Adjusting module, for obtain the information of at least one other user node by cooperation perception, and dynamically adjusts the priority of the replica server of described user node according to the information of described other user nodes.

As technique scheme preferably, user perspective definition module is determined the user perspective definition of each replica server to user node by following formula:

$F_{n_i^c}=\underset{j\∈N\backslash \{i,c\}}{\mathrm{\Σ}}\frac{d(n_i,n_c)d(n_j,n_c)}{d{(n_i,n_j)}^2}$

Wherein,

N _i, n _jrepresent respectively different replica servers;

N _cfor user node;

D (n _i, n _c) be user node n _cwith replica server n _ibetween network delay; D (n _j, n _c) be user node n _cwith replica server n _jbetween network delay; D (n _i, n _j) be two internodal network delays of replica server; N is the replica server sum in User, N={n _i| i ∈ { 0 ... k}}.

As technique scheme preferably, wherein said priority block is determined the priority of each replica server that each user node is corresponding according to user perspective definition F, wherein the priority of higher this replica server of user perspective definition F of replica server is higher.

As technique scheme preferably, described adjusting module obtains the priority of replica server of at least one other user node or the best copy server that described other user nodes are selected by cooperation perception, dynamically to adjust the priority of replica server of described user node.

As technique scheme preferably, adjusting module specifically comprises:

Probability calculation unit, for according to the visual angle definition F of each replica server of User, can calculate user node n _iselect replica server s _jprobability P _ij

$P_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\∈\{\mathrm{1,2}\·\·\·n\}}{\overset{n}{\mathrm{\Σ}}}{F^i}_{S_j}}$

Wherein n _iin the User of individual user node, comprise N replica server s ₁, s ₂, s ₃s _n;

Cooperation perception unit, for obtain the priority of the replica server of at least one other user node by cooperation perception, or user node obtains by cooperation perception the best copy server that at least one other user node is selected;

Probability adjustment unit, dynamic for the information obtaining according to cooperation perception unit, obtain the user node n after adjustment _iselect replica server s _jprobability p _ir', as this replica server s _jpriority.

The beneficial effect of technique scheme of the present invention is as follows:

In the embodiment of the present invention, first user node can calculate this user node and select replica server s _jprobability P _ij, and this replica server s _jrespond the probability of this user node, then by the selected best copy server of other user nodes s _j, adjust this replica server s _jrespond the probability q of first user node described in this _ij, and come again dynamically to adjust described first user node with this and select this replica server s _jprobability P _ij', and then adjust the probability of all replica servers that first user node is corresponding.Therefore the embodiment of the present invention can improve the effect of user node selection best copy server, prevents network congestion and server overload.

Accompanying drawing explanation

Fig. 1 is the schematic network structure that existing CDN-P2P merges content distributing network;

Fig. 2 is the method flow schematic diagram of the embodiment of the present invention;

Fig. 3 is an example of User.

Embodiment

For making the technical problem to be solved in the present invention, technical scheme and advantage clearer, be described in detail below in conjunction with the accompanying drawings and the specific embodiments.

The embodiment of the present invention proposed a kind of best copy server selection method that utilizes User cooperation perception, and its flow process as shown in Figure 2, comprising:

Step 1, set up User for each user node, and user node in definite User all replica servers that can connect.

Concrete, as shown in Figure 3 be an example of User, in this User, expressed all replica servers that each user node can connect.Wherein, the replica server setup parameter thresholding that can connect user node in the time determining User, obviously can not be as the replica server of best copy server to discharge.

Step 2, determine the user perspective definition of each replica server that each user node in User is corresponding.

Concrete, can with each user node to the network delay of each replica server of correspondence as this replica server the user perspective definition to this user node.The network delay of user and replica server is less, and in this User, the user perspective definition to this replica server is higher.Concrete, can determine the user perspective definition of each replica server to user node by following formula:

$F_{n_i^c}=\underset{j\∈N\backslash \{i,c\}}{\mathrm{\Σ}}\frac{d(n_i,n_c)d(n_j,n_c)}{d{(n_i,n_j)}^2}$ (formula 1)

Wherein,

N _i, n _jrepresent respectively different replica servers;

N _cfor user node;

D (n _i, n _c) be user node n _cwith replica server n _ibetween network delay; D (n _j, n _c) be user node n _cwith replica server n _jbetween network delay; D (n _i, n _j) be two internodal network delays of replica server; N is the replica server sum in User, N={n _i| i ∈ { 0 ... k}}.

Can determine the user perspective definition of each replica server in the User of each user node according to above-mentioned formula (1), and determine replica server according to user perspective definition; Because User definition is determined according to network delay, therefore can be determined at easily the nearer replica server of user distance and respond user's request.

Step 3, determine the priority of each replica server that each user node is corresponding according to user perspective definition.

Concrete, in the invention process, the priority of the replica server that user perspective definition F is larger is higher.The priority that is to say the replica server less with the network delay of user node is higher.

Step 4, obtain the information of at least one other user node by cooperation perception, and dynamically adjust the priority of the replica server of described user node according to the information of described other user nodes.

The perception that wherein cooperates refers to that group collaboration environment can, in the situation that not affecting other node cooperative works, pass to each cooperative node by the information notice of a node, makes each cooperative node can perceive each other the other side's state information.A replica server can be multiple users' contiguous replica server, and multiple users can, from same replica server Gains resources, can cause the problems such as Mean Time of Systemic Response is long, and server load is unbalanced, network congestion.User node can obtain the solicited message of other user nodes by cooperation perception.Therefore user node can dynamically select/change the replica server of request response as the best in the case of the disposal ability of consideration replica server and/or load condition, can dynamically change the contiguous replica server that will ask response, selection disposal ability and/or the good replica server of load condition are as best copy server.

Wherein, the information of other described user nodes can be the priority of the replica server of at least one other user node, or the best copy server of described other user nodes selections.That is to say, what this user node will obtain is the replica server that other user nodes will send request.

First user node can calculate this user node and select replica server s _jprobability P _ij, and this replica server s _jrespond the probability of this user node, then by the selected best copy server of other user nodes s _j, adjust this replica server s _jrespond the probability q of first user node described in this _ij, and come again dynamically to adjust described first user node with this and select this replica server s _jprobability P _ij', and then adjust the probability of all replica servers that first user node is corresponding.Therefore the embodiment of the present invention can improve the effect of user node selection best copy server, prevents network congestion and server overload.

Concrete, according to the visual angle definition F of each replica server in User, can calculate user node n _iselect replica server s _jprobability P _ij

$P_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\∈\{\mathrm{1,2}\·\·\·n\}}{\overset{n}{\mathrm{\Σ}}}{F^i}_{S_j}}$ (formula 2).

Due to this probability P _ijaccording to visual angle, definition F calculates, therefore larger this probability P of known visual angle definition F _ijalso just larger.

Wherein, in network, there is M user, and n _iin individual User, comprise N contiguous replica server s ₁, s ₂, s ₃s _n.

Meanwhile, because the user node in content distributing network (CDN) has reached some, therefore in network, there is the individual user of N (N≤M) simultaneously to a replica server s _jsend request information, replica server s _jresponse user n _ithe probability of request is q _ij

$q_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\∈\{\mathrm{1,2}\·\·\·n\}}{\overset{n}{\mathrm{\Σ}}}{F^i}_{S_j}}$ (formula 3).

In the time that multiple users' whiles are to same replica server solicited message in network, may cause the problems such as average response time is long, system load is uneven, network congestion.In embodiments of the present invention, user node can be according to the disposal ability of replica server and/or load condition, obtain the state information of user node by the cooperation perception between user, and the priority of the replica server to user node self is adjusted according to above-mentioned parameter, adjust user and select the probability P of contiguous replica server _ij.

Its principle is: just foregoing, in system, there is N user simultaneously to a replica server s _jsolicited message, user n _ifrom User, select replica server s _jprobability drop to:

${P_{\mathrm{ij}}}^{\′}=P_{\mathrm{ij}}\·q_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\∈\{\mathrm{1,2}\·\·\·n\}}{\overset{n}{\mathrm{\Σ}}}{F^i}_{S_j}}\·\frac{{F^i}_{S_j}}{\underset{i\∈\{\mathrm{1,2}\·\·\·n\}}{\overset{n}{\mathrm{\Σ}}}{F^i}_{S_j}}$ (formula 4).

Therefore can be by adjusting the selecteed probability P of replica server _ijselect with the priority of adjusting replica server, to avoid the problem such as laod unbalance and network congestion.

Due to user node n _iuser in the selected probability sum Σ P of all replica servers _i=1, therefore user n _iselect replica server s _jprobability decline time, select the probability of other N-1 replica server to increase accordingly.

Because the selected probability of each replica server is proportional, after adjusting, its probability also can proportionally increase.That is:

$\left\{\begin{array}{c}\mathrm{\&Sigma;}{P_i}^{\&prime;}=1\\ \frac{{p_{\mathrm{ir}-1}}^{\&prime;}}{{p_{\mathrm{ir}}}^{\&prime;}}=\frac{F_{\mathrm{ir}-1}}{F_{\mathrm{ir}}}(r\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\left\}\right)\\ {p_{\mathrm{ir}}}^{\&prime;}<1\end{array}\right\}$ (formula 5)

Wherein p _ir' be the selected probability of replica server after adjusting.

As can be seen here, user can constantly adjust and select replica server s according to cooperation perception _jprobability to reset the priority of replica server, to select suitable replica server as best copy server.

Meanwhile, the embodiment of the present invention has also proposed a kind of best copy server selective system of utilizing User cooperation perception, comprising:

User module, is used to each user node to set up User, and user node in definite User all replica servers that can connect;

User perspective definition module, for the user perspective definition of each replica server corresponding to each user node of definite User;

Priority block, for determining the priority of each replica server that each user node is corresponding according to user perspective definition;

Adjusting module, for obtain the information of at least one other user node by cooperation perception, and dynamically adjusts the priority of the replica server of described user node according to the information of described other user nodes.

As technique scheme preferably, user perspective definition module is determined the user perspective definition of each replica server to user node by following formula:

$F_{n_i^c}=\underset{j\&Element;N\backslash \{i,c\}}{\mathrm{\&Sigma;}}\frac{d(n_i,n_c)d(n_j,n_c)}{d{(n_i,n_j)}^2}$

Wherein,

N _i, n _jrepresent respectively different replica servers;

N _cfor user node;

D (n _i, n _c) be user node n _cwith replica server n _ibetween network delay; D (n _j, n _c) be user node n _cwith replica server n _jbetween network delay; D (n _i, n _j) be two internodal network delays of replica server; N is the replica server sum in User, N={n _i| i ∈ { 0 ... k}}.

Wherein said priority block is determined the priority of each replica server that each user node is corresponding according to user perspective definition F, wherein the priority of higher this replica server of user perspective definition F of replica server is higher.

Wherein, the information of other described user nodes can be the priority of the replica server of at least one other user node, or the best copy server of described other user nodes selections.That is to say, what this user node will obtain is the replica server that other user nodes will send request.

Like this, first user node can calculate this user node and select replica server s _jprobability P _ij, and this replica server s _jrespond the probability q of this user node _ij, then by the selected best copy server of the second user node s _j, adjust this replica server s _jrespond the probability of first user node described in this, and readjust the described first user node of dynamic adjustment with this and select this replica server s _jprobability, and adjust the probability of all replica servers that first user node is corresponding with this.

That is: described adjusting module obtains the priority of replica server of at least one other user node or the best copy server that described other user nodes are selected by cooperation perception, dynamically to adjust the priority of replica server of described user node.

Wherein, adjusting module specifically comprises:

Probability calculation unit, for according to the visual angle definition F of each replica server of User, can calculate user node n _iselect replica server s _jprobability P _ij

$P_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\}}{\overset{n}{\mathrm{\&Sigma;}}}{F^i}_{S_j}}$

Wherein n _iin the User of individual user node, comprise N replica server s ₁, s ₂, s ₃s _n;

Cooperation perception unit, for obtain the priority of the replica server of at least one other user node by cooperation perception, or user node obtains by cooperation perception the best copy server that at least one other user node is selected;

Probability adjustment unit, dynamic for the information obtaining according to cooperation perception unit, obtain the user node n after adjustment _iselect replica server s _jprobability p _ir', as this replica server s _jpriority.

Wherein, the operation principle of described adjusting module is:

Cooperation perception refers to that group collaboration environment can, in the situation that not affecting other node cooperative works, pass to each cooperative node by the information notice of a node, makes each cooperative node can perceive each other the other side's state information.A replica server can be multiple users' contiguous replica server, and multiple users can, from same replica server Gains resources, can cause the problems such as Mean Time of Systemic Response is long, and server load is unbalanced, network congestion.User node can obtain the solicited message of other user nodes by cooperation perception.Therefore user node can dynamically select/change the replica server of request response as the best in the case of the disposal ability of consideration replica server and/or load condition, can dynamically change the contiguous replica server that will ask response, selection disposal ability and/or the good replica server of load condition are as best copy server.

Concrete, according to the visual angle definition F of each replica server in User, can calculate user node n _iselect replica server s _jprobability P _ij

$P_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\}}{\overset{n}{\mathrm{\&Sigma;}}}{F^i}_{S_j}}$ (formula 2).

Due to this probability P _ijaccording to visual angle, definition F calculates, therefore larger this probability P of known visual angle definition F _ijalso just larger.

Wherein, in network, there is M user, and n _iin individual User, comprise N contiguous replica server s ₁, s ₂, s ₃s _n.

Meanwhile, because the user node in content distributing network (CDN) has reached some, therefore in network, there is the individual user of N (N≤M) simultaneously to a replica server s _jsend request information, server s _jresponse user n _ithe probability of request is

$q_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\}}{\overset{n}{\mathrm{\&Sigma;}}}{F^i}_{S_j}}$ (formula 3).

In the time that multiple users' whiles are to same replica server solicited message in network, may cause the problems such as average response time is long, system load is uneven, network congestion.In embodiments of the present invention, user node can be according to the disposal ability of replica server and/or load condition, obtain the state information of user node by the cooperation perception between user, and the priority of the replica server to user node self is adjusted according to above-mentioned parameter, adjust user and select the probability P of contiguous replica server _ij.

Its principle is: just foregoing, in system, there is N user simultaneously to a replica server s _jsolicited message, user n _ifrom User, select replica server s _jprobability drop to:

${P_{\mathrm{ij}}}^{\&prime;}=P_{\mathrm{ij}}\&CenterDot;q_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\}}{\overset{n}{\mathrm{\&Sigma;}}}{F^i}_{S_j}}\&CenterDot;\frac{{F^i}_{S_j}}{\underset{i\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\}}{\overset{n}{\mathrm{\&Sigma;}}}{F^i}_{S_j}}$ (formula 4).

Therefore can be by adjusting the selecteed probability P of replica server _ijselect with the priority of adjusting replica server, to avoid the problem such as laod unbalance and network congestion.

Due to user node n _iuser in the selected probability sum Σ P of all replica servers _i=1, therefore user n _iselect replica server s _jprobability decline time, select the probability of other N-1 replica server to increase accordingly.

Because the selected probability of each replica server is proportional, after adjusting, its probability also can proportionally increase.That is:

$\left\{\begin{array}{c}\mathrm{\&Sigma;}{P_i}^{\&prime;}=1\\ \frac{{p_{\mathrm{ir}-1}}^{\&prime;}}{{p_{\mathrm{ir}}}^{\&prime;}}=\frac{F_{\mathrm{ir}-1}}{F_{\mathrm{ir}}}(r\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\left\}\right)\\ {p_{\mathrm{ir}}}^{\&prime;}<1\end{array}\right\}$ (formula 5)

Wherein p _ir' be the selected probability of replica server after adjusting.

As can be seen here, user can constantly adjust and select replica server s according to cooperation perception _jprobability to reset the priority of replica server, to select suitable replica server as best copy server.

Can be found out for illustrating of the embodiment of the present invention by above-mentioned, in the time that user node is selected replica server, can know by the cooperation perception between user node the information of the replica server of other user node requests, dynamically adjust the replica server priority of self, to determine that dynamically best replica server is as best copy server.Can establish best copy server by the cooperation perception between user node like this, reduce Mean Time of Systemic Response, avoid the problem of replica server laod unbalance, optimize network performance.

The above is the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, do not departing under the prerequisite of principle of the present invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.

Claims (10)

1. a best copy server selection method that utilizes User cooperation perception, is characterized in that, comprising:

Step 1, set up User for each user node, and user node in definite User all replica servers that can connect;

Step 2, determine the user perspective definition of each replica server that each user node in User is corresponding;

Step 3, determine the priority of each replica server that each user node is corresponding according to user perspective definition;

Step 4, obtain the information of at least one other user node by cooperation perception, and dynamically adjust the priority of the replica server of described user node according to the information of described other user nodes.

2. the best copy server selection method that utilizes User cooperation perception according to claim 1, is characterized in that, described step 2 is specially:

Determine the user perspective definition of each replica server to user node by following formula:

$F_{n_i^c}=\underset{j\&Element;N\backslash \{i,c\}}{\mathrm{\&Sigma;}}\frac{d(n_i,n_c)d(n_j,n_c)}{d{(n_i,n_j)}^2}$

Wherein,

N _i, n _jrepresent respectively different replica servers;

N _cfor user node;

D (n _i, n _c) be user node n _cwith replica server n _ibetween network delay; D (n _j, n _c) be user node n _cwith replica server n _jbetween network delay; D (ni, n _j) be two internodal network delays of replica server; N is the replica server sum in User, N={n _i| i ∈ { 0 ... k}}.

3. the best copy server selection method that utilizes User cooperation perception according to claim 2, it is characterized in that, described step 3 is specially: determine the priority of each replica server that each user node is corresponding according to user perspective definition F, wherein the priority of higher this replica server of user perspective definition F of replica server is higher.

4. according to the best copy server selection method that utilizes User cooperation perception described in claim 1 or 2 or 3, it is characterized in that, described step 4 is specially:

User node can calculate the probability of this user node selection replica server, and this replica server responds the probability of this user node; By the selected best copy server of other user nodes, adjust the probability that this replica server responds first user node described in this, and readjust described first user node with this and select the probability of this replica server, and adjust the probability of all replica servers that first user node is corresponding.

5. the best copy server selection method that utilizes User cooperation perception according to claim 4, is characterized in that, described step 4 is specially:

Step 41, according to the visual angle definition F of each replica server in User, can calculate user node n _iselect replica server s _jprobability P _ij

$P_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\}}{\overset{n}{\mathrm{\&Sigma;}}}{F^i}_{S_j}}$

Wherein n _iin the User of individual user node, comprise N replica server s ₁, s ₂, s ₃s _n;

Step 42, user node obtain the priority of the replica server of at least one other user node by cooperation perception, or user node obtains by cooperation perception the best copy server that at least one other user node is selected;

The information that step 43, described user node obtain according to step 42 is dynamic, obtains the user node n after adjustment _iselect replica server s _jprobability p _ir', as this replica server s _jpriority.

6. a best copy server selective system of utilizing User cooperation perception, is characterized in that, comprising:

User module, is used to each user node to set up User, and user node in definite User all replica servers that can connect;

User perspective definition module, for the user perspective definition of each replica server corresponding to each user node of definite User;

Priority block, for determining the priority of each replica server that each user node is corresponding according to user perspective definition;

Adjusting module, for obtain the information of at least one other user node by cooperation perception, and dynamically adjusts the priority of the replica server of described user node according to the information of described other user nodes.

7. the best copy server selective system of utilizing User cooperation perception of stating according to claim 6, is characterized in that, user perspective definition module is determined the user perspective definition of each replica server to user node by following formula:

$F_{n_i^c}=\underset{j\&Element;N\backslash \{i,c\}}{\mathrm{\&Sigma;}}\frac{d(n_i,n_c)d(n_j,n_c)}{d{(n_i,n_j)}^2}$

Wherein,

N _i, n _jrepresent respectively different replica servers;

N _cfor user node;

D (n _i, n _c) be user node n _cwith replica server n _ibetween network delay; D (n _j, n _c) be user node n _cwith replica server n _jbetween network delay; D (n _i, n _j) be two internodal network delays of replica server; N is the replica server sum in User, N={n _i| i ∈ { 0 ... k}}.

8. the best copy server selective system of utilizing User cooperation perception of stating according to claim 7, it is characterized in that, wherein said priority block is determined the priority of each replica server that each user node is corresponding according to user perspective definition F, wherein the priority of higher this replica server of user perspective definition F of replica server is higher.

9. according to the best copy server selective system of utilizing User cooperation perception described in claim 6 or 7 or 8, it is characterized in that, described adjusting module obtains the priority of replica server of at least one other user node or the best copy server that described other user nodes are selected by cooperation perception, dynamically to adjust the priority of replica server of described user node.

10. the best copy server selection method that utilizes User cooperation perception according to claim 9, is characterized in that, adjusting module specifically comprises:

Probability calculation unit, for according to the visual angle definition F of each replica server of User, can calculate user node n _iselect replica server s _jprobability P _ij

$P_{\mathrm{ij}}=\frac{{F^i}_{S_j}}{\underset{j\&Element;\{\mathrm{1,2}\&CenterDot;\&CenterDot;\&CenterDot;n\}}{\overset{n}{\mathrm{\&Sigma;}}}{F^i}_{S_j}}$

Wherein n _iin the User of individual user node, comprise N replica server s ₁, s ₂, s ₃s _n;

Cooperation perception unit, for obtain the priority of the replica server of at least one other user node by cooperation perception, or user node obtains by cooperation perception the best copy server that at least one other user node is selected;

Probability adjustment unit, dynamic for the information obtaining according to cooperation perception unit, obtain the user node n after adjustment _iselect replica server s _jprobability p _ir', as this replica server s _jpriority.

Images