Energy-aware Web Service Composition Technology Research

With the development of Web service technology, the industry and academia both regardservice composition as one of the important backbone technology. Meanwhile, since the Webservice is widely used, the large scale composed service has emerged across the Internet. As aresult, the energy consumption of large scale composed service is also increasing rapidly,which has a bad influence to the overall performance of service composition. In order to makebetter use of the service composition and to improve the system efficiency, the research ofenergy aware service composition is no time to delay. The traditional theory of energyconsumption optimization is not very suitable for service composition area, it needs to befurther expanded and improved. Up to now, the research of energy consumption optimizationis still infrequent in the area of Web service composition, there’re plenty of problems to besolved urgently. At the big background of green computing, the research aims at lowering theenergy consumption of Web service composition during its running time. At the preconditionof satisfying user’s global QoS constraints, we’ll achieve the goal of energy aware servicecomposition. During composing services, reducing energy consumption during the servicecomposition process can not only improve the efficiency of service composition, but alsoraise the utilization of system resource, so an energy-aware service composition method isneeded.From the perspective of energy consumption, model the energy consumption of singleservice firstly. Then, the work builds energy-aware model for composed service,and it appliesto the situation of which one user request has multiple feasible paths. After that, according tothe given power model, this work gives the definition to the non-functional attributes, i.e. theQoS constraints, and at the same time, to get the corresponding constraint equations based onstate-space model. When finishing the whole model, the main context come to solve andcalculate the model problem, and also give the two calculation conditions, i.e. the power ratioand the QoS constraints. The work divides the calculation process into two parts: one isfinding the feasible path and the other is distributing the request rate. To the former, theprocess is so complicated that the text leaves a whole chapter alone to talk about the specificsearch algorithm; as to the latter, the author uses the chaos theory to calculate, which theobjective function is the total energy consumption of service composition and the independentvariable is the request rate. At the part of finding the feasible path, the author proposes aservice composition method based on planning graph to solve the problem with its concept ofhierarchy. The algorithm can be divided into three steps, as the following: hierarchical search, best QoS calculation and feasible path selecting.At last, the work verifies the feasibility and effectiveness of this method through a groupof simulation experiments. Compared with other traditional methods, the proposed methodcan not only guarantee the users’ requests, but also cut down the total energy consumption ofcomposed service effectively.