QoS Based Optimal Resource Allocation in Service Composition for Heterogeneous Devices

Service-Oriented Computing(SOC) rapidly gains importance in the industry to built complex web service applications. The web service plays a very important role in SOC. Service composition is the process of adding services to the existing one using functional attributes of multiple services rapidly and effectively. One of the critical issues in the composition is non-consideration of Quality of service (QoS) parameters which can lead to inaccurate and low performing composite service. Next to selection tasks in the composition process is to allocate resource to selected services in composite service is also crucial. Resource allocation in the composition is the process of allocating component service to selected candidate service for each and every candidate services in composite service. In the cloud environment, component service holds physical computational resource or infrastructural resource. This paper, present mapping of resource allocation problem in composition to 0/1 Multi-dimensional knapsack problem which is NP-hard computational problem. The work in the paper provides the solution to the resource allocation problem with the help of a dynamic programming strategy. The computational simulation was carried out and revealed that the dynamic programming strategy as not suitable for a large number of services considered during the allocation process. Further development of the solutions for such problems should be attempted by applying various heuristics and meta-heuristics approaches. However, after analysis of these approaches, the work suggests to study and apply meta-heuristics to address this type of problems for an optimal solution with the feasibility.