Page - 363 - in Intelligent Environments 2019 - Workshop Proceedings of the 15th International Conference on Intelligent Environments

For example, Table 2 andTable 3 showhow theAHPmethod is applied to obtain weight for theQoScriteria:Cost,Response time(Rt),Reliability (Rel), andThroughput (Th). Table2. Comparaisonmatrix Cost Rt Rel Th Cost 1 0,3333 0,2 0,1429 Rt 3 1 0,5 0,1667 Rel 5 2 1 0,25 Th 7 6 4 1 Sum 16 9,3333 5,7 1,5595 Table3. Weights criteria Weights Cost 0,063 0,036 0,035 0,092 0,056 Rt 0,188 0,107 0,088 0,107 0,122 Rel 0,313 0,214 0,175 0,16 0,216 Th 0,438 0,643 0,702 0,641 0,606 Checksum 0,563 0,357 0,298 0,359 1 5. WebservicesrankingwithTOPSISmethod The TOPSIS is amethodwhose purpose is to be able to classify in order of choice a certainnumberof alternativeson thebasisof a set of favorable andunfavorable criteria. It is part of the techniques used in the field of multi-criteria decision-making. It was developed byHwang andYoon in 1981 as an alternative to theELECTREmethod and canbeconsideredasoneof itsmostwidelyacceptedvariants. The ranking of alternatives in TOPSIS is based on “similarity to the ideal solu- tion”,which avoids having the same resemblance to the positive ideal and the negative ideal[20][21].Thedifferent stepsof theTOPSISmethodaredescribed in [17]. 6. Experiment In this sectionwe show howwe apply our approach for on a real example.We show how to select the best service among a set of real services related to SMS sending web services. The used similar services are as follows: SMSWS,SMS,SendMessages, SendSMS,andSendSMSWorld.Eachwebservice is characterizedby four criteria: reli- ability(Rel), throughput(Th),Cost and response time(Rt).The reliabilityand throughput are favorablecriteria and the twoother criteria arenot favorable. Theassignedvalues for thewebservicecriteriaareused tocreateadecisionmatrix. Forourexample theobtainedmatrix ispresentedas follows: H.Belouaaretal. /AHPandTOPSISMethods forWebServiceSelection 363