CN101364899A - Web QoS evaluating method based on cloud model - Google Patents

Abstract

A quality of service (Qos) quantitative evaluation method for web service applied in the Internet environment is mainly used for solving the problem of Qos evaluation based on user view in a computation paradigm of SOC (Service-Oriented Computing). The method comprises the following steps: (1) quantitatively processing and evaluating the uncertainty of subjective data according to historical subjective evaluation data for web service; (2) quantitatively processing and evaluating the objective Qos of the web service according to the deviation degree of the objective Qos of the web service and SLA (Service Level Agreement); and (3) carrying out comprehensive Qos evaluation of the web service on the basis of the quantized data of the steps (1) and step (2). The method can solve the problem of uncertainty for subjective quantitative evaluation of Qos and the problem of subjective-objective comprehensive evaluation of Qos, thereby improving the quality of discovery, selection and combination of the web service in the dynamic and open Internet environment.

Description

Web QoS evaluating method based on cloud model

Technical field

The present invention relates under the internet environment, SOC calculates the evaluation method of web service quality in the pattern, particularly, relates to a kind of based on the master of cloud model, the integrated evaluating method of objective service quality.

Background technology

Along with the development of web service technology, increasing web service and ISP appear on the Internet.Although existing web service standard, SOAP for example, UDDI, WSDL can support the description of serving, and finds and binding, individual or mercantile customer must correctly choose web service and supplier's thereof challenge in the face of how in open internet environment.Owing in numerous candidate web services, may have a plurality of web services that similar or identical function are provided.Therefore, service quality (Qos) then becomes the extra benchmark that careful differentiation is carried out in service to web.

Traditional Q oS is almost contained complete specification requirement, performance for example, successful execution rate, reliability, availability, fail safe etc.But, in internet environment, be not enough only from realizing hierarchy of skill definition and investigating service quality.Because the opening of the Internet makes terminal use's participation ability obtain unprecedented lifting.The user can estimate used service once, can propose expectation to concrete service quality details before using service.Thereby from the service consumer view, the subjective credit worthiness of web service and defer to the degree of service level agreement (SLA) becomes the emphasis that the numerous researchers in this field pay close attention to.

The Web service credit worthiness is meant the trusting degree of service consumer to web service and ISP, is determined the height of its credit worthiness usually by consumer's subjective Quality of experience.SLA is service consumer and the service level agreements that the supplier reaches before to be that web service is actual call, and is made up of some objective QoS attributes usually, as the response time of serving, the successful execution rate, availability, reliability etc., the objective Qos attribute among the SLA is called the parameter of SLA again.Yet for the consumer, the web service has different implications with the credit worthiness of service providers.For example, have the web service of high credit worthiness, show that this service of consumer confidence satisfies the ability of function and not function requirement.On the other hand, for the ISP, its credit worthiness shows whether the consumer believes that it has the favorable service intention.

In ecommerce and Web Community field, already trust problem in opening, the isomery internet environment is carried out comparatively deep research.In these fields, trusting relationship is regarded as than the security mechanism more essential a kind of security relationship of tradition based on ' mandate-authentication '.The researcher thinks that the grade evaluation of trusting depends on user's existing experience, is based on the subjective judgement result of conviction, has tangible randomness and ambiguity characteristics.And adopt probability theory and Fuzzy Set Theory, respectively the randomness and the ambiguity of trusting are carried out independent studies.

In the web service context, current more existing from the subjective view of user, the research work that QoS is studied.Based on the credit worthiness measure of subjective assessment, calculate the mean value of different credit worthiness opinion ratings, perhaps weighted average.Based on the measure of SLA, think simple by estimating the subjective marking of user, be not enough to catch the difference between the active service quality and the agreement grade of service.Thereby, proposing a kind of new credit worthiness tolerance and be called verity, verity weighs web service and ISP's credibility according to the difference between the active service quality and the agreement grade of service.

Yet still there is some limitation in the research work of relevant credit worthiness in the web service field.At first, these methods are not clearly distinguished web service and supplier's credit worthiness, and actual conditions usually are promptly to estimate the web service with the same metric index again ISP's credit worthiness to be weighed.Secondly, the basic skills of credit worthiness evaluation is, based on the method for the calculating mean value of frequency, even verity tolerance still is to adopt similar mean value calculation method.This method is not reasonably considered the randomness and the ambiguity of subjective trust.In order to overcome above-mentioned limitation, need a kind of rational tolerance mechanism, to distinguish web service and ISP's credit worthiness.And on this basis, provide a kind of can be rationally, effectively in conjunction with the method for quantitatively evaluating of randomness and ambiguity characteristics.Thereby the credit worthiness evaluation information that accurately, reasonably reflects web service and ISP finally reaches and improves the web service discovery, chooses the target with combination quality.

Summary of the invention

At the mode of service consumer participation service evaluation and service level agreement, the present invention introduces two new credit worthinesses tolerance, and promptly ability credit worthiness and intention credit worthiness are respectively applied for web service ability and ISP's service intention is estimated.

The ability of supposing is trusted space AT=[0,1] be a certain amount of domain, C represents certain qualitativing concept to web service ability trusting degree at being used on this domain.For arbitrary x ∈ AT, x is a random number with steady tendency to the degree of membership μ ∈ [0,1] of notion C.Then the distribution of x on AT is called ability trust cloud, is designated as ATC (x).

Ability is trusted the credit worthiness that has reflected the web service, and this credit worthiness mainly depends on service consumer the subjectivity of web service ability is experienced, and promptly the height of ability credit worthiness is calculated by the subjective marking value of user.Usually, subjective credit worthiness scoring interval can be real number value arbitrarily in [0, n].If n＜〉1, when the ability of carrying out credit worthiness is estimated, need this interval is mapped to [0,1] interval, calculate so that finish the ability credit worthiness with unified standard.

Suppose that intention trusts space IT=[-1,1] be a certain amount of domain, C represents the qualitativing concept of service consumer to the trusting degree of ISP's service intention and motivation on this domain.For arbitrary x ∈ IT, x is a random number with steady tendency to the degree of membership μ ∈ [0,1] of notion C.Then the distribution of x on IT is called ability trust cloud, is designated as ITC (x).

Intention is trusted the credit worthiness that cloud is used to measure the ISP.This tolerance index expression web serves the degree that QoS property value in the actual invoked procedure departs from SLA.Before SLA was service call, the service level agreement of reaching between service consumer and supplier showed the expectation of service consumer to service quality, and the ISP is to the promise of service quality.Usually SLA is made of one group of parameter, and each parameter can be considered a kind of water dust type.The actual value of each parameter and its desired value poor can be considered the quantized data of a class water dust.Because the negotiations process of SLA still needs consumer's participation, and, still need consumer's subjective judgment to the cognition of final departure degree.Therefore, still there are subjective cognitive randomness and ambiguity during the intention credit worthiness is estimated.For this reason, the present invention adopts cloud model that the randomness and the ambiguity of intention credit worthiness are unified to describe, and is rational.

On this basis, the present invention proposes a kind of credit worthiness based on cloud model and quantizes evaluation method, is used to solve web service discovery based on subjective trust, chooses and the combination quality problem.This method can be served the credit worthiness evaluation result of intention in conjunction with web service ability and ISP, solves the trust decisions problem that the web service was correctly found, chooses and made up to service consumer preferably.

The inventive method is specific as follows:

Step 1:

According to the definition of ability credit worthiness, designed capacity is trusted cloud, and the ability that comprises is trusted the type selecting of cloud, and general the selection has the one dimension normal state cloud of universality well, but does not get rid of the selection of using the other types cloud; The interval mapping mode of trusting the space to ability of subjective credit worthiness; Can set the qualitativing concept rank in case of necessity.

Step 2:

According to the design result of ability trust cloud, from the credit worthiness database, obtain the historical subjective credit worthiness evaluation information of web service.Use reverse cloud generating algorithm, capacitation is trusted three numerical characteristics of cloud, and structuring capacity is trusted characteristic vector.

Step 3: trust characteristic vector and credit worthiness quantification formula according to ability $\mathrm{RS}=\mathrm{Ex}\×e^{-\mathrm{He}}+\frac{b}{c}\mathrm{Ex}(c=b+1)---\left(1\right),$ Calculate the ability credit worthiness of web service.B and c are factor of influence, are used to limit the accuracy rating of credit worthiness.

Step 4: design idea is trusted cloud, comprises the type selecting of intention trust cloud, and selection principle is identical with ability trust cloud; Choosing of SLA parameter, for example the SLA parameter can comprise service response time, availability etc.; Can set the qualitativing concept rank in case of necessity.

Step 5: obtain the historical data of SLA parameter, and according to formula D _n=v _e-v _a(2) and D _p=v _a-v _e(3), a plurality of SLA parameters being carried out the data sheet tonality handles; Formula (2) can be identical monotonicity with the SLA improve parameter unification with different monotonicities with (3).

According to formula $v=\frac{c_i-c_{\min }}{c_{\max }-c_{\min }}---\left(4\right),$ The SLA parameter is carried out normalized, obtain being intended to trust the water dust data of cloud.According to reverse cloud generating algorithm, calculate three numerical characteristic values that intention is trusted cloud, make up intention and trust characteristic vector.

Step 6: calculate the intention credit worthiness according to formula (1).

Step 7: according to formula (5) and (6), calculate comprehensive credit worthiness, promptly trust scoring.

$\mathrm{TS}={({\mathrm{RS}}_A+1)}^{\mathrm{Threshold}({\mathrm{RS}}_A,{\mathrm{RS}}_I)}-1---\left(5\right)$

$\mathrm{Threshold}({\mathrm{RS}}_A,{\mathrm{RS}}_I)=\left\{\begin{array}{c}0,{\mathrm{RS}}_I\≤{\mathrm{\δ}}_I\\ 0,{\mathrm{RS}}_A\≤{\mathrm{\δ}}_A\\ {\mathrm{RS}}_I,\mathrm{others}\end{array}\right.---\left(6\right)$

Description of drawings

Fig. 1 prototype system structural representation

Fig. 2 trust evaluation schematic flow sheet

Embodiment

Understand and enforcement the present invention the existing embodiments of the invention of describing in conjunction with the accompanying drawings for ease of persons skilled in the art.

As shown in Figure 1, for implementing the prototype system structural representation of method described in the present invention, by at least one user terminal apparatus 1, at least one trust evaluation center 2, and at least one credit worthiness database 3 is formed.User terminal apparatus 1 can be any wired or wireless communication ability that has, and the device that can carry out information exchange with trust evaluation center 2, as PC, mobile phone etc.User terminal 1 is by any wire communication line 401, or wireless communication line 402, carries out alternately with trust evaluation center 2.Credit worthiness database 301 is used to store the historical subjective credit worthiness evaluation information of web service, SLA database 302 is used to store the SLA agreement, and the actual executory SLA supplemental characteristic of service, 301 and 302 can carry out data interaction with trust evaluation center 2 by any wire communication line 401 or order control link 403.

Trust evaluation center 2, by at least one master control set 201, at least one ability credit worthiness evaluating apparatus 202, at least one intention credit worthiness evaluating apparatus 203, at least one is trusted scoring apparatus 204 and forms.Master control set 201 is connected with device 202,203,204 by order control link 403, and the execution of responsible whole evaluation procedure and finish information interaction with user terminal apparatus 1.Device 202 is responsible for finishing the ability credit worthiness evaluation of trusting cloud based on ability, device 203 is responsible for finishing the intention credit worthiness evaluation of trusting cloud based on intention, device 204 is finished the synthetic of ability credit worthiness and intention credit worthiness, and generates the task of final trust appraisal result.

In the specific implementation process of the present invention, can dispose a plurality of distributed trust evaluation center 2 as required, and a plurality of trust evaluation center 2 can be carried out information and data interaction mutually.In addition, the physical centralization configuration can be carried out in trust evaluation center 2 and credit worthiness database 3, also can carry out decentralized configuration.

As shown in Figure 2, existing in conjunction with prototype system structural representation shown in Figure 1 for implementing the trust evaluation flow chart of method described in the present invention, concrete implementation step is described in detail as follows:

Step-:

User terminal apparatus 1 by 401 or 402 and trust evaluation center 2 connect.By and the information exchange of 201 of master control sets, selected one group of web service is as by evaluation object.According to being estimated mechanism and scoring system by the credit worthiness of evaluation object, master control set 201 is finished ability and is trusted the cloud design, the ability that comprises is trusted the type selecting of cloud, and general the selection has the one dimension normal state cloud of universality well, but does not get rid of the selection of using the other types cloud; The interval mapping mode of trusting the space to ability of subjective credit worthiness; Can set the qualitativing concept rank in case of necessity.

Step 2:

Master control set 201 obtains the historical credit worthiness data of web service from credit worthiness database 301.If desired, then the credit worthiness interval is mapped as standard capability and trusts the space.Then, according to reverse cloud generating algorithm, generative capacity is trusted three numerical characteristics of cloud.The ground that is without loss of generality, whether the present invention unifies not do the pressure requirement to different scorings, and different web services can be marked by different evaluation mechanism, but may influence the accuracy of evaluation.Therefore, in identical evaluation mechanism and scoring system, the effect of the method for the invention can be better.

According to the historical subjective assessment information of web service, use reverse cloud generating algorithm, computing capability is trusted three numerical characteristic values of cloud, and structuring capacity is trusted characteristic vector.Ability is trusted characteristic vector and is made of the numerical characteristic Ex and the He value of cloud model, and shape is as＜Ex, He 〉.According to this characteristic vector, when carrying out ability credit worthiness quantification evaluation, the physical significance of Ex and He is respectively: the representative value that Ex estimates as credit worthiness, the i.e. average credit worthiness level of credit worthiness; He represents the departure degree of credit worthiness and average credit worthiness, can think that He has embodied the stability of credit worthiness.When Ex is big, can think that credit worthiness is higher, on the contrary lower; When He hour, can think that credit worthiness is more stable, on the contrary its less stable.

Step 3:

Master control set 201 is to ability credit worthiness evaluating apparatus transmitting capacity credit worthiness computation requests.202 according to credit worthiness quantification formula $\mathrm{Rs}=\mathrm{Ex}\×e^{-\mathrm{He}}+\frac{b}{c}\mathrm{Ex}(c=b+1)---\left(1\right),$ Calculate the ability credit worthiness of web service, and ability credit worthiness result of calculation is back to 201.B and c are factor of influence, are used to limit the accuracy rating of credit worthiness.

Step 4:

Master control set is finished and is intended to trust the cloud design, comprises the type selecting of intention trust cloud, and selection principle is identical with ability trust cloud; Choosing of SLA parameter, for example the SLA parameter can comprise service response time, availability etc.; Can set the qualitativing concept rank in case of necessity.The present invention does not force to limit the number and the type of SLA parameter, can construct the parameter sets of SLA according to actual needs.

Step 5:

Master control set 201 obtains the historical data of SL Λ parameter, and according to formula D _n=v _e-v _a(2) and D _p=v _a-v _e(3), a plurality of SLA parameters being carried out the data sheet tonality handles.Because the monotonicity difference of different service quality, for example, service response time has monotone decline, and promptly influence time is big more, and its quality is poor more.And the monotonicity of monotonically increasing qualitative attribute in contrast, and as the successful execution rate, when successful implementation rate is big more, its quality is good more.D _nAnd D _vExpression has the water dust data of descending and incremental, V respectively _eAnd V _aThe desired value and the actual value of the expression water dust value that distributes.Therefore, by formula (2) and (3), can with SLA improve parameter unification identical monotonicity, i.e. monotonic increase with different monotonicities.

Master control set 201 is according to formula $v=\frac{c_i-c_{\min }}{c_{\max }-c_{\min }}---\left(4\right),$ The SLA parameter is carried out normalized, obtain being intended to trust the water dust data of cloud.V represents normalization result, c in the formula (4) _iRepresent i water dust data, c _MinAnd c _MaxExpression minimum respectively and maximum water dust value.According to reverse cloud generating algorithm, calculate three numerical characteristic values that intention is trusted cloud, make up intention and trust characteristic vector.The implication that intention is trusted characteristic vector is identical with ability trust characteristic vector.

Step 6:

Master control set 201 sends intention credit worthiness evaluation request to intention credit worthiness evaluating apparatus 203.203 calculate the intention credit worthiness according to formula (1), and will be intended to credit worthiness result of calculation and be back to 201.

Step 7:

Master control set 201 sends the trust evaluation computation requests to trusting scoring apparatus 204.204 according to formula formula (5) and (6), calculate comprehensive credit worthiness, promptly trust scoring.204 will trust appraisal result is back to 201, will finally trust appraisal result by 201 and be back to user terminal apparatus 1.

$\mathrm{TS}={({\mathrm{RS}}_A+1)}^{\mathrm{Threshold}({\mathrm{RS}}_A,{\mathrm{RS}}_I)}-1---\left(5\right)$

$\mathrm{Threshold}({\mathrm{RS}}_A,{\mathrm{RS}}_I)=\left\{\begin{array}{c}0,{\mathrm{RS}}_I\≤{\mathrm{\δ}}_I\\ 0,{\mathrm{RS}}_A\≤{\mathrm{\δ}}_A\\ {\mathrm{RS}}_I,\mathrm{others}\end{array}\right.---\left(6\right)$

RS _AAnd RS _IRepresent the ability credit worthiness of web service and ISP's intention credit worthiness respectively.TS represents final trust scoring, and formula (5) satisfies the constraints of following trust evaluation:

1. trust scoring with RS _AAnd RS _IIncrease and increase, reduce and reduce

2. has only the RS of working as _AAnd RS _IAll be 1 o'clock simultaneously, trust scoring TS and equal 1

3. work as RS _AOr RS _IWhen being lower than certain threshold value, trusting scoring TS and equal 0

4.RS _AValue big more, trust scoring with RS _IVariation and the speed that changes is fast more, vice versa

Formula (6) is the threshold calculations formula, is used to ensure constraints 3.δ _AAnd δ _IRepresent RS respectively _AAnd RS _IThreshold value, these threshold values can be specified by the user according to the application need of reality, perhaps calculate according to certain given principle.

Two the credit worthiness tolerance indexs (ability credit worthiness and intention credit worthiness) that relate in the inventive method have been avoided web service and the unsharp problem of ISP's credit worthiness evaluation index; In the mode that quantizes, for the web service consumer finish rationally, exactly the web service discovery, choose and combination provides guidance, the randomness and the ambiguity that have solved subjective trust evaluation in the web service quality are difficult to the unified problem of expressing.Trust characteristic vector based on cloud model has directly perceived, succinct quantification characteristics, has avoided the web service consumer in the face of multiple trust evaluation mechanism, reaches the decision error that is produced when multiple evaluation result represents form.

In the specific embodiment of the invention, be example, introduced the embodiment that implements the inventive method with the normal state cloud.Medium cloud types of models of the present invention, and the reverse cloud generating algorithm relevant with the cloud model type can be chosen according to actual needs; And in the invention process process, it is in full accord with the type of intention trust cloud that the ability that do not require is trusted cloud in principle.In addition, the present invention is trust evaluation formula (5) and (6) in order to trust the elementary sum rational faculty of scoring, has only set described four constraintss.Therefore, at cloud model type, relevant cloud operative algorithm, and the variation of aspect such as two kinds of cloud model type consistency selection, and the trust evaluation method that satisfies above-mentioned constraints or expansion constraints all belongs to the protection range of claim of the present invention.

Claims (8)

1. method for quantitatively evaluating based on the service quality (Qos) that is applied to the service of web under the internet environment of cloud model.It is characterized in that: (1) use ability is trusted the numerical characteristic of cloud, according to the subjective trust scoring with randomness, ambiguity characteristics, the service ability of web service is carried out quantitatively evaluating; (2) use intention to trust cloud, describe the departure degree of objective Qos of Web service and SLA (service level agreement), web is served objective Qos variation carrying out quantitatively evaluating; (3) on the basis of 1,2 quantized data, finish the master of web service, the overall merit of objective service quality.

2. a kind of method for quantitatively evaluating that is applied to the service quality (Qos) of web service under the internet environment based on cloud model according to claim 1 is characterized in that ability is trusted the quantitative credit worthiness evaluation of estimate that cloud employing ability trust space AT represents the web service in (1); The quantitative domain interval of AT is [0,1], and subjective credit worthiness then represents credit worthiness low more more near 0, more near 1, then represents credit worthiness high more.

3. a kind of method for quantitatively evaluating that is applied to the service quality (Qos) of web service under the internet environment according to claim 1 based on cloud model, it is characterized in that the trust of intention described in (2) cloud adopts intention to trust space IT and represents quantitatively to be intended to the credit worthiness evaluation, the quantitative domain of IT interval is [1,1], and approaching more-1, then represent credit worthiness low more,, then represent credit worthiness high more more near 1.

4. a kind of method for quantitatively evaluating that is applied to the service quality (Qos) of web service under the internet environment according to claim 1 based on cloud model, it is characterized in that (1) further comprises: ability is trusted the cloud design, ability is trusted the cloud numerical characteristic and is generated, structuring capacity is trusted characteristic vector＜Ex, He 〉.

5. a kind of method for quantitatively evaluating that is applied to the service quality (Qos) of web service under the internet environment according to claim 1 based on cloud model, it is characterized in that (2) further comprise: intention is trusted the cloud design, intention is trusted the cloud numerical characteristic and is generated, make up intention and trust characteristic vector＜Ex, He 〉.

6. a kind of method for quantitatively evaluating that is applied to the service quality (Qos) of web service under the internet environment according to claim 1 based on cloud model, it is characterized in that (1) further comprises: trust characteristic vector＜Ex according to ability, He 〉, quantize to calculate the ability credit worthiness of web service.

7. a kind of method for quantitatively evaluating that is applied to the service quality (Qos) of web service under the internet environment according to claim 1 based on cloud model, it is characterized in that (2) further comprise: trust characteristic vector＜Ex according to intention, He 〉, web ISP's intention credit worthiness is calculated in quantification.

8. a kind of method for quantitatively evaluating that is applied to the service quality (Qos) of web service under the internet environment based on cloud model according to claim 1 is characterized in that (3) may further comprise the steps:

1) ability credit worthiness and the threshold value that is intended to credit worthiness are set;

2) finish the synthetic of ability credit worthiness and intention credit worthiness according to threshold value, generate final trust scoring.

Images