CN102932460A - Campus network peer-to-peer (P2P) incentive method based on contribution values - Google Patents

Abstract

The invention relates to a campus network P2P incentive method based on contribution values. According to the method, current nodes serve as main design factors for resource sizes and sharing rates shared by other nodes, and nodes with more shared resources can obtain more contribution values; and the higher the sharing rates are, the more contribution values can be obtained. The method takes other factors such as single shared resource sizes, sharing time and resource classification into consideration simultaneously. According to the method, various factors are combined to serve as the basis of single node contribution values, accordingly, resource qualities and fairness, economy, differences and diversities of incentive mechanisms are improved.

Description

A kind of Campus Networks P2P motivational techniques based on contribution margin

Technical field

The present invention relates to a kind of Campus Networks P2P motivational techniques based on contribution margin, be applied to technical field of the computer network.

Background technology

Peer-to-peer network (P2P) is a kind of opening that is made of the computer of some cooperations, anonymous network.From the viewpoint of service, all nodes all are reciprocity in this network, and each node is the supplier of service, also are the recipients of service.Traditional P2P system does not provide effective incentive mechanism, and each node is not subordinated to mutually in the network, is self-management and autonomous decision resources contribution, thereby makes most of nodes all show selfish behavior, and then cause a large amount of existence of following two kinds of phenomenons:

(1) (Free Riders) hitchhikes ^[1]Phenomenon: most of nodes all are reluctant to share the resource of oneself, but the free available resources of using other nodes to provide are provided with a free rider's identity.

(2) public tragedy (Tragedy of Common) phenomenon: because the propagation of a large amount of same asset on the P2P network system, thereby having caused for a long time, shared resource reaches unanimity gradually, nobody is ready to share new resource, repeats in a large number resource and is used without limit by most of network nodes as the public resource of nonexcludability.

In the Gnutella statistics, 70% user never shares any file, and 50% file polling responds the sharing users from 1%.The inventor herein is by finding that to the data analysis of the campus network P2P of unit one belongs to data interchange platform 50% node is never shared any file, provided by 8% node during 90% shared file.The selfish behavior of node has had a strong impact on the balance of P2P system resource thus, has reduced the efficient of P2P network ^[5], therefore, must effectively suppress with public tragedy behavior hitchhiking.

The existence of the selfish behavior of node has had a strong impact on the realization of P2P system original intention.At present, the scholar has proposed multiple solution to this problem, wherein effect better, more representational class methods are to take incentive mechanism, mainly comprises incentive mechanism based on prestige, based on the incentive mechanism of mutual reciprocity and mutual benefit model with based on incentive mechanism of ideal money model etc.

(1) based on the incentive mechanism of prestige

Mechanism based on prestige mainly is the concept of having introduced a grade in the P2P network, and namely each node is estimated the credit value that draws with its other contiguous node in obtaining by network according to own historical behavior situation at network.Emphasis based on the incentive mechanism of prestige is to attempt QoS and the node of node are combined to the degree that system makes contributions.

The behavior of other node of node maintenance is historical, and is using these information to make a strategic decision afterwards.Node A determines whether to provide service to Node B, and the situation of service is provided based on Node B node in the system.Based on the motivational techniques of prestige to large-scale node, highly dynamic member consists of and the P2P system of non-frequently repeated interaction is with good expansibility.The difference of various incentive mechanisms based on prestige is that the calculating of credit value and prestige are to the mapping of concrete decision strategy.

In the calculating of credit value, depend on the recommendation that other node provides based on the incentive mechanism of prestige, therefore need to successfully manage the dishonest recommendation that other node provides.Generally there are three kinds of methods to realize the calculating of credit value: 1) by privately owned history, namely to utilize direct experience to carry out trust evaluation.This method is applicable to the smaller situation of system scale, and is difficult to practicality in the asymmetric situation of interest between the node.2) share objective prestige (Objective Reputation) based on global history.Whether this method can't be distinguished the experience that other node provides genuine and believable, can't tackle the attack of dishonest feedback, and be difficult to obtain the global history of node in non-structured P2P network, therefore is difficult to realize.3) the subjective prestige (Subjective Reputation) of sharing based on global history.This method faces the problem that is difficult to realize equally in non-structured P2P network.For this reason, method is arranged according to the characteristics of non-structural P 2 P network, propose based on the historical subjective prestige of sharing of part.This method can be tackled dishonest feedback problem effectively, in the destructuring network, is easy to simultaneously realize, and be a kind of feasible scheme.The method uses trust information exchange limited between the neighbor node to realize trust evaluation, and introduces the tactic behavior change that adaptive forgetting factor is dealt with malicious node.In the calculating of recommendation trust degree, we use service trust and recommendation trust degree to calculate the method that is separated, and dynamically recommend the renewal of confidence level according to interaction results, can effectively tackle in partnership deceptive practices and the libel action of malicious node.In order to prevent that node from no longer providing service after setting up good prestige, the method is considered time factor when trust value calculates, and trusts in time decay when not having new interaction experiences to occur.Therefore, the calculating of trust value is the function of time correlation.

(2) based on the incentive mechanism of mutual reciprocity and mutual benefit model

That service providing node in the P2P network can obtain after providing service for other nodes that certain is directly preferential based on the basic thought of directly reciprocal mechanism, this mechanism is the very effective mode of Resource Exchange, but it only exists for the effective weakness of the transmission of certain Resource Exchange.That is to say that each node exists only in the once exchange for the maintenance of other node historical informations, after this time exchange finishes, provides mutually the node of service will know nothing contribution situation each other.Although eMule and eDonkey adopt the two-way energisation mode of transience that transmits for once, its considers that the historical record of peer node also carries out the integration queuing, thereby difference service is provided.But show also only having the uploading of 3% node can be to the node of being familiar with before own in most of situation through statistics, and other is uploaded difference service can't be provided.

(3) based on game theoretic incentive mechanism

Tit-for-tat is typically based on game theoretic incentive mechanism, and this mechanism is used for a Critical policies of game theoretic prisoner game, can be according to the dominating stragegy of adversary's policy selection oneself.This tactful core is exactly in order to maximize the interests of oneself, can to work in coordination between the participant.According to the rate of uploading of node contribution, with N node from N to 1 descending sort.Upload node and will select u node to go to upload, u is the concurrent linking number of uploading of node, so node N, N-1 ..., N-u+1 will be selected and download shared file.Owing to upload node and obtained the higher rate of uploading from an above u node, therefore upload node will be connected u node maintenance connection.Consider now node N-u and since its contribution to upload rate less, it can not obtain to download connection, the download rate of node N-u is zero.In order to obtain the download rate in next round, node N-u has to improve the rate of uploading of oneself, will exceed at least the rate of uploading that node N-u+1 provides, and guarantee obtains to download in next round and connects.So repeatedly game, each node is just had to contribute and ownly maximum is uploaded the broadband to system for obtaining maximum interests.

(4) based on the incentive mechanism of ideal money

Main thought based on the incentive mechanism of ideal money is that Resource consumers should be the certain remuneration of resource provider payment, and node exchanges various services for this ideal money.Adopt this machine-processed system will design the smooth circulation that a whole set of perfect economic model is controlled ideal money, need simultaneously system that effective safety measure is provided, guarantee fairness.This incentive mechanism model more complicated implements relatively difficulty.

From the angle of incentive mechanism design, system should satisfy fairness, economy, otherness and multifarious characteristics.The contribution that is node is more, and the income of its acquisition is more; The resource that provides should be provided, avoid the wasting of resources; Stress encourage original shared node, accelerate the renewal of resource; Encourage sharing of unexpected winner resource, guarantee the abundant of resource.On this angle, contrasted above-mentioned three kinds of incentive mechanisms, comparing result sees Table 1.As can be seen from Table 1, above-mentioned three kinds of methods respectively have deficiency, need to propose a kind of new method, improve this mechanism.

Four kinds of incentive mechanism contrasts of table 1

Algorithm	Fairness	Economy	Otherness	Diversity
					Prestige	Generally	Generally	Generally	Generally
Mutual reciprocity and mutual benefit	Good	Good	Generally	Generally
					Game theory	Good	Good	Generally	Generally
Ideal money	Generally	Poor	Generally	Generally

Summary of the invention

For above the deficiencies in the prior art, the present invention proposes certainly to realize at the P2P network software method of flow optimization with it.Share energetically own resource with excitation P2P network system node, solve and hitchhike in the Campus Networks and public tragedy problem, the raising network service quality promotes the normal operation of whole P2P system.

Purpose of the present invention is achieved through the following technical solutions:

A kind of Campus Networks P2P motivational techniques based on contribution margin, the method comprises the steps:

1) obtains shared cycle of shared file in the node, the shared cycle of file

$G_{\mathrm{Ti}}=a_T-b_T\×e^{-\frac{\mathrm{Ti}}{c_T}};$

Obtain the size of shared file in the node, the size of shared file is:

$G_{\mathrm{Si}}=\left\{\begin{array}{cc}{a}_G\&times;S_i& S_0\&GreaterEqual;S_i\\ \mathrm{Si}& S_0<S_i<S_1\\ b_G\&times;S_i& S_i\&GreaterEqual;S_1\end{array}\right.$

Obtain and share nodes in the node, the shared nodes of file is:

$G_{\mathrm{Ui}}=a_U+e^{-\frac{\mathrm{Ui}-1}{b_U}}$

Obtain the contribution margin weight of shared file in the node, the contribution margin weight is:

$G_{\mathrm{Vi}}=a_V\&times;{V_i}^{b_V}$

Inverse ratio according to the occupation rate of shared file arranges weight; Concrete formula is:

$\mathrm{Ci}=\frac{1}{K_i}$

According to time of network utilization is arranged different weights to the contribution margin of different periods, improve the weight of idle period, time of every day is divided into the different periods, calculate the weight Pi of i period, the weight Pi=a of i period _P* Li;

2) obtain the contribution margin that shared file produces in the said system, contribution margin is:

Ai=Pi×Ci×Di×G _Si×G _Ui×G _Ti×G _Vi (7)

Wherein Di is shared type,

Obtain the contribution margin of all shared files of node:

$A=\underset{i}{\mathrm{\&Sigma;}}\mathrm{Ai}---\left(8\right)$

The contribution margin upper limit a that the user obtains in the unit of account time ₀, obtain the total contribution margin of node:

$B=a_0(1-e^{{-b}_0\&times;A})$

3) according to the contribution margin of node i to platform, calculate its contribution margin note and be B, arbitrary node is arranged the determined threshold D of a contribution margin, as B less than D the time, its authority is limited;

4) when node fails to reach the desired contribution margin of system, then system only allows it to upload and provide the shared resource operation, until the node contribution margin reaches system requirements;

5) to different contribution margin users, arrange and download the restrictions such as number different every days, improve the user and share enthusiasm.

The invention has the advantages that:

(1) in the identical duration of contrast, this motivational techniques of the present invention have improved the enthusiasm of nodes sharing resource, have increased system's available resources total amount.

(2) by this motivational techniques of the present invention, more node has participated in sharing of resource, the resource of in the past being shared by individual node, and beginning is by increasing nodes sharing.Be embodied in generally the increase of sharing nodes, be embodied on the details that to share nodes be that 1 file begins to reduce, in the multinode direction set.

(3) by this motivational techniques of the present invention, the proportion of C1 class file has dropped to 68% by 81%, and the file of C2 class has risen to 21% by 12%, and other class files also have been increased to 11% by 7%.

(4) by this motivational techniques of the present invention, restrain to median the life cycle of file, namely has the short period file of some to prolong shared time of file, with conforming to of expection.

Description of drawings

Fig. 1: shared file distribution map life cycle;

Fig. 2: shared file size distribution figure;

Fig. 3: G _SiFunction curve diagram;

Fig. 4: shared file nodes distribution map;

Fig. 5: shared file category distribution figure;

Fig. 6: shared file nodes distribution map;

Fig. 7: shared file category distribution figure;

Fig. 8: shared file distribution map life cycle.

Embodiment

Each node will provide service to whole network according to initial agreement in the P2P network, comprises sharing memory space and spendable resource, also can enjoy memory space and the resource that other nodes provide in the network simultaneously.Difference between the different incentive mechanisms is mainly reflected in utility function, node, and to enjoy service ability and node be the aspects such as system's contextual definition of having contributed, measurement point selection.When the contribution margin algorithm design, encourage as far as possible the enthusiasm of node, the assurance system moves continually and steadily.

It is that resource size and the shared rate that other nodes are shared is main design factor that the present invention takes present node, and the node that shared resource is more will obtain more contribution margins; Shared rate is higher also will to obtain more contribution margins.The present invention also considers other a plurality of factors simultaneously.Single shared resource size: single shared resource is larger, and the contribution margin of acquisition is larger, and on the contrary, single shared resource is less, and the contribution margin of acquisition is less, and purpose is to encourage the high-quality resource of nodes sharing; Share time length: the shared time is longer, and the contribution margin of acquisition is larger, and the shared time is shorter, and the contribution margin of acquisition is less; Resource classification: for different resource classification, give different to share it, excitation nodes sharing scarce resource.The comprehensive various factors of the present invention improves the fairness of resource quality and incentive mechanism as the foundation of individual node contribution margin with this, and is economical, otherness and diversity.

Incentive mechanism based on contribution margin mainly is that node is served discriminatively.Namely contribute large node to obtain quality and serve preferably, correspondingly, contribute less node can only obtain second-rate service in addition can not get the service.The below is to realizing with it certainly that at the P2P network software method of flow optimization describes in detail.

Incentive mechanism algorithm based on contribution margin

When the contribution margin algorithm design, encourage as far as possible the enthusiasm of node, the assurance system moves continually and steadily.To simplify the analysis, it is that file size and the shared rate that other nodes are shared is main design factor that this paper takes present node, consider simultaneously single shared file size, share the factors such as time length, document classification, comprehensive various factors is as the foundation of individual node contribution margin.For convenient narration, introduce 7 descriptive definitions, see Table 2.

The definition of using in the table 2 contribution margin algorithm

Definition	Describe
		Definition 1	I represents the file i in the P2P system.
Definition 2	Ti represents the life cycle (TTL) of file i.
		Definition 3	Ui represents the shared nodes of file i.

Definition 4	Si represents file i size.
		Definition 5	Ci represents the classification of file i.
Definition 6	Di represents file i type identification.
		Definition 7	Pi represents that file i shares rate.

The present invention is based on following hypothesis:

1) infrastructure such as P2P platform user, equipment are sound, have certain scale.

2) influencing factor for contribution margin is: share time length, shared file size, share nodes, share speed, share kind, share period, file type.

3) for two shared file i and j, if Ti〉Tj, then the contribution margin G of i file generation _TiContribution margin G greater than the generation of j file _Tj

4) for any two shared file i and j, if Si〉Sj, then the contribution margin G of i file generation _SiContribution margin G greater than the generation of j file _Sj

5) for any two shared file i and j, if Ui〉Uj, then the contribution margin G of i file generation _UiContribution margin G less than the generation of j file _UjSimultaneously, the contribution margin summation of file i generation is greater than the contribution margin summation of file j generation.

6) for any two shared file i and j, if Vi〉Vj, then the contribution margin G of i file generation _ViContribution margin G in the generation of j file _Vj

7) initiatively provide the contribution margin of file-sharing greater than the contribution margin of sharing file.

Central idea of the present invention is for to serve discriminatively node.Namely contribute large node to obtain quality and serve preferably, correspondingly, contribute less node can only obtain second-rate service in addition can not get the service.Detailed process is:

The establishment of contribution margin computing formula

Step 1: the file-sharing cycle is longer, and it is more to obtain contribution margin, thereby improves the continuity of resource.Consider the ageing of seed, when the file-sharing cycle was longer, contribution margin began to tend to be steady.Therefore use formula (1) as the file i Contribution Function of life cycle.From what Fig. 1 was not difficult to find out shared file individual minimum arranged within [10,12] week, and after in the long period file number basicly stable.The characteristic distributions that design parameter is shared the cycle according to platform file is debugged.

$G_{\mathrm{Ti}}=a_T-b_T\&times;e^{-\frac{\mathrm{Ti}}{c_T}}---\left(1\right)$

In the formula: G _TiRepresent the file i contribution margin of life cycle

a _TAnd b _TBe parameter to be determined, Ti represents the life cycle (TTL) of file i,

Step 2: the file of nodes sharing is larger, and it is larger then to obtain contribution margin.For obtaining the parameter of rule 2 regulations, distribution situation such as Fig. 2 of shared file size have been added up.For the diversity of file is provided, small documents is suitably improved the contribution margin weight, encourage sharing of super large file, the contribution margin weight of large file is provided, therefore adopt piecewise function to describe, G _SiThe curvilinear figure of function as shown in Figure 3, concrete formula is:

$G_{\mathrm{Si}}=\left\{\begin{array}{cc}{a}_G\&times;S_i& S_0\&GreaterEqual;S_i\\ \mathrm{Si}& S_0<S_i<S_1\\ b_G\&times;S_i& S_i\&GreaterEqual;S_1\end{array}\right.---\left(2\right)$

In the formula: G _SiThe contribution margin of list file i size, a _gAnd b _gBe parameter to be determined, Si represents file i size. step 3: the node contribution margin that same file is shared beginning is high, and along with the increase of sharing nodes, contribution margin reduces, and total contribution margin increases along with the increase of node.This rule effectively solves free-riding behavior, avoids simultaneously the waste of storage resources.The distribution situation of file-sharing nodes, as shown in Figure 4.Based on this, use formula (3) to come description document to share nodes to the impact of contribution margin.

$G_{\mathrm{Ui}}=a_U+e^{-\frac{\mathrm{Ui}-1}{b_U}}---\left(3\right)$

In the formula: G _UiThe contribution margin of the shared nodes of expression file i, a _UAnd b _UBe parameter to be determined, Ui represents the shared nodes of file i.

Step 4: for the interchange of accelerating resource with share, give and high contribution margin weight for shared file at a high speed, suitably reduce the weight of low speed shared file.Concrete formula is:

$G_{\mathrm{Vi}}=a_V\&times;{V_i}^{b_V}---\left(4\right)$

Work as b _VFor greater than 1 real number the time, can meet the demands.Compare as parametric description with shared rate, can better reflect details.In the formula: G _ViList file i shares the contribution margin of speed, a _VAnd b _VBe parameter to be determined, Vi represents the shared speed of file i.

Step 5: weigh from the angle of Campus Network characteristics and P2P system, for the diversity of affluent resources, be provided with shared file classification Ci in design during based on the incentive mechanism algorithm of contribution margin, encourage nodes sharing to be worth higher file.For shared relatively less study, the resources such as physical culture, we arrange weight according to the inverse ratio of the occupation rate of shared file; Concrete formula is:

$\mathrm{Ci}=\frac{1}{K_i}---\left(5\right)$

Ci represents the classification contribution margin weight of file i, K in the formula _iBe the shared percentage of i class file.

In addition for fear of public tragedy.Arrange simultaneously and share type Di, distinguishing provides resource and the proportion of sharing resource, encourages node that new resources are provided, and satisfies D1:D2=n2:n1.N1 is for initiatively providing the number of resource in the formula in a period of time, and n2 is the number of sharing resource in a period of time.

Step 6: consider the regular strong of school's daily schedule, avoid blocking up of peak period, we arrange different weights to the contribution margin of different periods, improve the weight of idle period, with this effect of dredging the peak even avoiding.The time of every day is divided into the different periods, and the size of the weight Pi of the i period therewith device free rate Li of period is directly proportional.Namely

Pi=a _P×Li （6）

Pi is the contribution margin weight of i period in the formula, and Li is the device free rate, a _PFor to be determined.

Step 7: the contribution margin that the File i of system produces is decided to be approximately according to above analysis:

Ai=Pi×Ci×Di×G _Si×G _Ui×G _Ti×G _Vi (7)

The impact on contribution margin of node All Files is decided to be approximately:

$A=\underset{i}{\mathrm{\&Sigma;}}\mathrm{Ai}---\left(8\right)$

For the ease of statistics, we set the contribution margin upper limit a that the user obtains in the unit interval ₀, utilize formula (9) to realize, the total contribution margin of node is designated as:

$B=a_0(1-e^{{-b}_0\&times;A})---\left(9\right)$

A in the formula ₀And b ₀Be parameter to be determined.

Differentiated services take contribution margin as standard

Step 1: for the node i of new adding, in the incipient stage, give certain download authority, after the time limit, calculate its contribution margin note and be B.A determined threshold D is set, as B less than D the time, its authority is limited.The distribution situation with the platform contribution margin is depended in the setting of parameter D.According to formula (9), calculate the contribution margin of all nodes, and carry out statistical analysis, according to the distribution situation of contribution margin, corresponding determined threshold D is set.

Step 2: when node failed to reach the desired contribution margin of system, then system only allowed it to upload and provide the shared resource operation, until the node contribution margin reaches system requirements.

Step 3: for different contribution margin users, arrange and download the restrictions such as number different every days, improve the user and share enthusiasm.

In the design based on the incentive mechanism model of contribution margin, the contribution margin of node is the basis of whole model, is the foundation that system carries out difference service, and difference service is the core of whole incentive mechanism model, is the key point of this model.

In order to assess the effect of new contribution margin computing formula and corresponding incentive mechanism, the P2P file-sharing platform in the Campus Networks that image data is used when utilizing parameter to analyze is tested, and this network is logined the independent IP of this system every day above 10000.After system has moved 5 months, adopt the data of new regulation front and back from the following aspects comparative analysis.

(1) in the identical duration of contrast, be incorporated herein the system of algorithm and do not introduce the P2P system of incentive mechanism, the available resources total amount is along with the situation of change of time.By introducing new rule, improved the enthusiasm of nodes sharing resource, increased system's available resources total amount.

(2) contrasted the change situation of introducing new regulation front and back system shared file nodes.After introducing new regulation, the distribution situation of shared file nodes as shown in Figure 6.Comparison diagram 4 is not difficult to find that more node has participated in sharing of resource, the resource of in the past being shared by individual node, and beginning is by increasing nodes sharing.Be embodied in generally the increase of sharing nodes, be embodied on the details that to share nodes be that 1 file begins to reduce, in the multinode direction set.

(3) situation of change of shared file kind in the system of introducing new regulation front and back.After introducing new regulation, comparison diagram 5 and Fig. 7 are not difficult to find that the proportion of C1 class file has dropped to 68% by 81%, and the file of C2 class has risen to 21% by 12%, and other class files also have been increased to 11% by 7%.

(4) introduce after the new regulation variation of life cycle of shared file in the system, can find that restrain to median the life cycle of file, namely have the short period file of some to prolong shared time of file, with conforming to of expection, as shown in Figure 8.

The present invention is directed to " hitchhiking " and " public tragedy " phenomenon that Campus Networks P2P system file extensively exists in sharing, analyzing on the basis that relatively has incentive mechanism now, proposed a kind of incentive mechanism based on contribution margin.The contribution margin that this mechanism utilizes the contribution margin function to carry out node calculates, and impels node for realizing that himself maximum revenue is to whole network dedicate resources.Improve the enthusiasm of node dedicate resources, improved fairness and the validity of system.Show by the experiment test in the campus network: this mechanism can effectively realize storage resources distributional equity in the P2P system, reaches the purpose that the excitation node participates in dedicate resources, suppresses the selfish behavior of node.

Should be appreciated that the above detailed description of technical scheme of the present invention being carried out by preferred embodiment is illustrative and not restrictive.Those of ordinary skill in the art is reading on the basis of specification of the present invention and can make amendment to the technical scheme that each embodiment puts down in writing, and perhaps part technical characterictic wherein is equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution break away from the spirit and scope of various embodiments of the present invention technical scheme.

Claims (1)

1. the Campus Networks P2P motivational techniques based on contribution margin is characterized in that the method comprises the steps:

1) obtains shared cycle of shared file in the node, the shared cycle of file

$G_{\mathrm{Ti}}=a_T-b_T\&times;e^{-\frac{\mathrm{Ti}}{c_T}};$

Obtain the size of shared file in the node, the size of shared file is:

$G_{\mathrm{Si}}=\left\{\begin{array}{cc}{a}_G\&times;S_i& S_0\&GreaterEqual;S_i\\ \mathrm{Si}& S_0<S_i<S_1\\ b_G\&times;S_i& S_i\&GreaterEqual;S_1\end{array}\right.$

Obtain and share nodes in the node, the shared nodes of file is:

$G_{\mathrm{Ui}}=a_U+e^{-\frac{\mathrm{Ui}-1}{b_U}}$

Obtain the contribution margin weight of shared file in the node, the contribution margin weight is:

$G_{\mathrm{Vi}}=a_V\&times;{V_i}^{b_V}$

Inverse ratio according to the occupation rate of shared file arranges weight; Concrete formula is:

$\mathrm{Ci}=\frac{1}{K_i}$

According to time of network utilization is arranged different weights to the contribution margin of different periods, improve the weight of idle period, time of every day is divided into the different periods, calculate the weight Pi of i period, the weight Pi=a of i period _P* Li;

2) obtain the contribution margin that shared file produces in the said system, contribution margin is:

Ai=Pi×Ci×Di×G _Si×G _Ui×G _Ti×G _Vi (7)

Wherein Di is shared type,

Obtain the contribution margin of all shared files of node:

$A=\underset{i}{\mathrm{\&Sigma;}}\mathrm{Ai}$

The contribution margin upper limit a that the user obtains in the unit of account time ₀, obtain the total contribution margin of node:

$B=a_0(1-e^{{-b}_0\&times;A})$

3) according to the contribution margin of node i to platform, calculate its contribution margin note and be B, arbitrary node is arranged the determined threshold D of a contribution margin, as B less than D the time, its authority is limited;

4) when node fails to reach the desired contribution margin of system, then system only allows it to upload and provide the shared resource operation, until the node contribution margin reaches system requirements;

5) to different contribution margin users, arrange and download the restrictions such as number different every days, improve the user and share enthusiasm.

Images