Based On The Service Quality Of Web Services Model And Applied Research

With the widespread proliferation of e-Business, the Internet service model transformed from data communication and information browsing to electronic trading and services, which bring Web Services to be widely employed. The e-Business applications demand dif-ferent dealing priority according to the users or services, thus provide differentiated service based on profit. Web services inherits best-effort mechanism which equally processing all the incoming requests and adopts FIFO (First In First Out) scheduling strategy, which the improved performance obtained from network QoS (Quality of Service) mechanism, e.g.IntServ/DiffServ etc., is seriously damaged. And it also cannot meet the varying QoS re-quirements of users. As the infrastructure is complex, there are many factors to influence the QoS of Web services. Changes in traffic patterns, denial-of-service attacks and the effects of infrastructure failures, low performance of Web protocols, and security issues over the Web create a need for Internet QoS standards. Often, unresolved QoS issues cause critical trans-actional applications to suffer from unacceptable levels of performance degradation, and it is skeptical if Web Services can be employed to key application. Thus it is urgent to re-searching QoS of Web Services. The investigating and employing QoS have some achieve-ment on different layers, which shows that it is feasible to fully manage QoS. This disserta-tion has the following innovation.We propose a consistent QoS computing model of Web Services. The current re-searches do not form a unit of QoS model and some QoS models are given on the different layers, e.g., single Web Service, Web service transaction and workflow. Those models are independent, which cause overloading the Web services provider and conflicts when apply-ing the models. The consistent model can unify the QoS management in an integrated struc-ture. The demonstrations also show that it improves the QoS.On the basis of consistent QoS computing model, we propose a QoS model of single Web Services including factors, computing method and strategy. The factors involve per-formance, reliability, cost, SLA and reputation. This model fully considers and satisfies the profit and requirement of users and providers. The strategy introduces self-adaptive sched-uling algorithm for priority buffers which guarantees that lower priority requirement do not influence the higher. The algorithm regards QoS value as weight of priority. Thus the higher...