QoS Based Web Services
Pradeep Tomar and Gurjit Kaur
Web Services are the implementation of client and server architecture in which various client and server applications communicate by exchanging data over hypertext transfer protocol over the web. The means of interoperable software applications over various platforms are web services. Web services are distinguished by their interoperable and extensible nature, with their descriptions, using XML. For attaining complex operations loosely coupled web services are used. To achieve values added services, simple web services interact to each other by exchanging data. Various open standards like Simple Object Access Protocol, Web Service Description Language, XML and Universal Description and Discovery Integration are used to integrate web applications through the Web services. Different standards have different roles, the tagging of data is done using XML, transmission of data is done using simple object access protocol, web services are described using description language and listing of available services is done trough UDDI. Various application from differing sources can communicate in real time through the web services, since XML is used for communicating. There is no restriction on using any specific operating platform or language for functioning of web services.
Role of Functional and Non- Functional attributes Web services are gaining interest with every second, thus it is becoming very important to differentiate between the services that perform the same kind of functionality. For this purpose functional and non-functional attributes play an important role. The functional and non functional attributes help the users to decide whether to go for the service or not. Several web services provide the relatively similar kind of functionality but have different QoS parameters involved. In order to provide the user with suitable service it is required to understand the need of the user. The QoS parameters are scenario based so are entitled to change in different cases. Since multiple QoS attributes be required at a time by the user, it is important to take composition of these QoS for selection of the web service. • Reliability Reliability refers to the ability to perform intended operations under some stated conditions with available resources. The web services should be reliable to provide the most hassle free experience to the end user. • Robustness A robust web service should perform consistent even when partially fed with ambiguous input. Web services should have a high degree of consistency. • Availability Considering immediate usage of the service, the service should be always be live or running. The service should always respond to every valid request of the user. The availability is the probability that the system is up and related to reliability. • Rating This QoS attribute is based on user or website submitted usage-statistical data. This attribute will help in knowing which service is most frequently used or well known among the users. • Service This attribute provides data regarding response quality to request. An important consideration will be quality of service with respect to cost of the web service selected. Better the service according to the price, better will be the rating for that very service. • Cost This attribute refers to the charges of using the service. It is an important factor in deciding whether to select the service or not. The better the service with equate availability and lower cost of service, will benefit the service provider with good rating. • Interoperability Interoperability refers to the ability of being operable at different environments so that the programmers don't have to worry for services to be written in any specific language or for any specific platform. Web services should be able to operate on changing platforms. • Security Service providers enforce enhanced security for web services by achieving protocols like confidentiality, authenticity, data integrity, encryption.
FUZZY APPROACH FOR WEB SERVICES A fuzzy logic is a computing technique which is widely used in various fields. It was developed by LotfiZadeh in the 1960s and 70s to model those problems in which uncertainty factor is involved. Fuzzy logic is a good extension of ordinary logic, where the main advantage is that we use fuzzy sets for the membership of a variable. Fuzzy logic can give many advantages over the ordinary logic.
Fuzzy Inference System The traditional idealistic mathematical approach has been improved to accommodate partial truth by the Introduction of fuzzy set theory by Professor LotfiA.Zadeh, [12], in1965. Fuzzy logic provides a convenient way to represent Linguistic variables and subjective probability. The motivation and justification for fuzzy logic is that the linguistic characterizations are less specific than the numerical ones. Most situations in the world require crisp actions. These actions are arrived by processing fuzzy information in figure. Fuzzy logic is used to provide means of inferring the fuzzy information to produce crisp actions. Fuzzy logic provides the tools to: Fuzzification: Transform world information from crisp to fuzzy information. Inference: Infer the fuzzy information to come to a fuzzy action. Composition: Aggregation of the outputs of all the fuzzy actions. Defuzzification: Transform back the fuzzy action to a crisp action.
Fuzzificaton Fuzzification is the process of making the crisp quality “fuzzy”. This allows addressing uncertainty in any parameter due to imprecision, ambiguity or vagueness. In artificial intelligence, the most common way to represent human knowledge is in terms of natural language i.e. linguistic variables. Depending upon the data and uncertainty, the inputs and the output parameters are fuzzified in terms of linguistic descriptors such as high, low, medium, and small to translate them into fuzzy variables e.g. fuzzy boundaries are parameters “age ”can be formed by the linguistic expressions such as “young”, “middle aged”, and “old”. Therefore, fuzzy sets for the inputs parameters and the required single output parameter are formulated based on the expert’s knowledge and experience in the particular domain.
Inference Having specified the expected number of faults and its influencing parameters, the logical next step is to specify how the expected numbers of faults vary as a function of the influencing parameters. Experts provides fuzzy rules in the forms of if..then statements that relate expected number of faults to various levels of influencing parameters based on their knowledge and experience. Fuzzy processor uses linguistic rules to determine what control action should occur in response to a given set of input values. Rules evolution also referred as fuzzy inference, evaluates each rule with the inputs that were generated from the fuzzification process.
Composition The inputs are combined logically using the AND operator to produce output response values for all expected inputs. The active conclusions are the combined into a logical sum for each membership function. A firing strength for each output membership function is computed. The fuzzy outputs for all rules are finally aggregated to one fuzzy set for various levels of consequence.
Defuzzification The logical sums are combined in the defuzzification process to produce the crisp output. To obtain a crisp decision from fuzzy output, the fuzzy set, or the set of singletons have to be defuzzified. There are several heuristic methods (defuzzification methods), like bisector method or the centroid method . One of them is e.g. to take the center of gravity as shown in the figure below. For the discrete case with singletons usually the mean of maximum method is used where the point with maximum singleton is chosen.
REVIEW OF LITERATURE OF QOS BASED WEB SERVICES N. Hema priya [1] proposed a architecture for the Web Services, Fuzzy Rule based algorithm. The Architecture contains of a Client, Service providers and their Registry as shown is fig.2 , which are considered as the main building elements for web services. There are number of service providers with different services, they get their service level agreements registered. Then the user searches for services according to their needs. Based on the type of services, some providers to provide authentication. Various QoS attributes are considered and their composition is also taken (if needed), using the proposed Web Service Architecture.
The architecture contained of a Fuzzy service discovery broker that acts as a middleware, a Fuzzy Engine, a Fuzzy Classifier for evaluating the QoS criteria for the registered service. Fuzzy Engine uses repository stored Inference Rules and gives a weight to each service. Various standards are used to implement web services, these standards play a different role in the whole service architecture. A service provider uses the internet to publish the service, Web Service Description Language handles the description whereas Universal Description and Discovery Integration keeps track of the files, and Simple Object Access Protocol is used to invoke the service whenever requested. Shuping Ran [2] discussed that web services have gained interest but the adoption hasn't kept pace and investigates that QoS parameters is one of the main reason for the slow adoption. The paper proposed a model for discovery of Web services in which various functional and quality of service attributes were used for discovering the service. Previous generation web services discovery models are highly unregulated because of the UDDI registries. A very high percentage of UDDI registries have unusable links. A new regulated model is proposed which can exist with unregulated registries. The purpose of unregulated registries was to offer services to those for whom service quality is of no significance. The applicants who needed service quality assurance were served using the regulated proposed model. Doru Todinca [3] proposed an approach in which user preferences and QoS characteristics play major role in selection of the services. The approach consisted of a vocabulary of description of services and their domain, and user preference based service selection that is handles by a trader, employing impaired comparison of services and algorithms for their ranking. The idea of this approach was that from using user preferences , automated fuzzy rules are used in fuzzy inference system for ranking the service. The paper presented the trials to estimate their approach consisting of prototype enactment of a service broker. A new prototype implementation of a service broker is devised, and fuzzy inference rules are used to solve the problem of selecting most apt match for each request, according to the QoS requirements, from a pool of imperfect services. This approach proposed a method to accomplish the QoS information using fuzzy categories, way to describe choices and requests in the service broker tool. This approach was capable of generating automated fuzzy rule for each set of individual choices or preferences. After generation of fuzzy rules, each candidate form the service pool was tested against these rules. Masri and Mahmoud [4] proposed an idea to solve problems using keyword search mechanism and Web service Relevance Function that assigning, measuring relevancy of the service. A crawler engine was employed to provide quality ranking. The users could use this model for searching and managing criteria based on their preferences. The service with highest rating is considered the most relevant as per interests. This approach reduced the cost of service. In this paper, for searching apt web services a blend of web service attributes were taken as constraints. This allowed to extend the web services repository building architecture by handing out quality driven findings of web services. This approach showcases the effectiveness of employing QoS attributes in search requests , outputting results as constraints and elements. To install confidence among users before calling for a service, proper information and service assurance was provided about that service by employing QoS attributes while finding web services by preferences. A service ranking mechanism was also proposed. Shen and Su [5] regarded a new model for web services based on automata and formal logic. To represent the semantic properties on service behaviors a new query based language was developed. Chengying Mao [6] adopted a method which used Petri Nets for web service composition to compute the complexity. The method provided two metric sets for evaluating the composition, logic of execution and dependencies in the workflow, as in a business operation depiction. This paper proposed some metrics for complexity of Web services workflow described by Petri-Net. Initially Petri-Net representations and their corresponding basic elements of business process were analyzed. Following that, control flow metrics were devised. Later in this paper the main aspects of web service composition were addressed. The two types metrics were proposed for workflow of web service composition, which are count-based metrics and execution path-based metrics. Data mining technique (WEKA) is used in Susila and Vadivel (2011) [7] presented such a scheme, which used Web Service Discovery Language files to choose the most apt service . This paper proposed extended Service Oriented Architecture which used data mining technique over QoS attributes to discover suitable web services. QoS attributes are taken into consideration like Availability, Security, Latency, Cost, then WEKA algorithm is applied to the data set. It provides tools for classification of data, clustering of data, regression, pre-processing and visualization. Palanikkumar [8] used Bee Colony Optimization Metaheuristic for optimizing the QoS locally ,which is based on evolution. To solve deterministic and combinatorial problems, BCO is employed in this approach. Qusay H. Mahmoud , Eyhab Al-Masri [9] advocated that the users shouldn't waste endless times to go through the UDDI based business registries to find suitable web services on mobile devices. The process of searching for these web services must be very effective and effortless. This paper discussed issues related to effective and time saving access and discovery of web services across several business registries. In this paper n new discovery engine was introduced named Web Service Crawler engine. The role of WSCE is to go through several registries and generating a central repository of web services which was used in faster and efficient discovery of the web services. This paper presented a new framework that was capable of extending the Web Service Repository Builder architecture by improving the discovery of web services without having to change current standards. This work introduced new crawler engine that was able to crawl across different available UBRs. Hoi Chan, Trieu Chieu [10] described a new method in which the web services were ranked and selected based on certain QoS attributes and prior knowledge. In this approach the web services QoS attributes were treated and web service relationships were targeted and represented as matrix. Singular value decomposition technique and adaptive weighting system were used to get the high order correlations among web services and their related QoS attributes to estimate the selection of recommended web services. A new approach was proposed for efficient selection and composition of web services using the SVD technique. Also the idea of Quality matrix for web services was introduced as a model for storing QoS information for every service which is being monitored. Glen Dobson, Stephen Hall [11] described a non-functional requirement ontology that was utilized for structuring and expressing constraints as a part of service quality specification. The ontology proposed was a part of European funded SeCSE integrated project. Liangzhao Zeng, Anne H.H. Ngu [1] proposed an loose, unbiased and potent model for QoS computation for selection of web services by implementing QoS registry in an unrealistic phone service for market place application. A framework was proposed in which the QoS model was extendible, QoS information was either given by service providers or computed from execution monitoring done by the end users or was gathered from the customers feedback considering the QoS criteria. This framework was aimed at enhancing the QoS modeling, computation. The proposed framework consisted of extensible QoS model, preference oriented service ranking and unbiased and open QoS computation. Qian MA, Hao WANG [13] proposed a framework for Semantic web services which was semantic QoS aware, which was achieved by formulation of semantic matchmaking and constraint programming together. Firstly, to define quality of service data was presented in service descriptions using QoS ontology. After this, syntactic matchmaking was changed into semantic way by employing ontology reasoning. When it was confirmed that various concepts are compatible, complex Quality of service conditions were solved and a algorithm for selection was proposed to provide optimal deal. Lijun Mei, W.K. Chan [14] presented an adaptable framework that prohibited problem causing external services from being used in service based application in a company. The framework used effectual web service description language information in public registries to estimate a schema of the network of the services. Along this, link analysis was done on the schema to recognize the services that were popular among different service consumers at any particular instant. Service composition was procedurally made using the services that were highly popular. The framework was capable of recognizing reliable external services for use in service based applications within organizational atmosphere. In this framework, a consumer X using a service Y, was capable of counting the number of times Y served X, and the number of cases of failure. Daniel A. Menasce' , Vinod Dubey [15] extended the previous work that was done on QoS brokerage for service oriented architecture. In this paper a service selection QoS broker was designed, implemented and evaluated, that increased a utility function for consumers of the service. The purpose of utility function was to enable the stakeholders to assign value to the system considering it as a function of various attributes like efficiency, availability and response time. This work assumed that the users will provide their utility functions along with the cost preference on the requested service, to the QoS broker. The implementation of broker and the service was done on java enterprise edition platform. This work also addressed the performance and availability issues of the QoS broker and was extended to provide a flexible and loosely coupled integration scheme. This paper also presented ideas by developing components and services using java enterprise edition platform.
