A Diffraction Service Composition Approach Based on S-ABCPC: An Improved ABC Algorithm

1. Introduction

With the development of service-oriented computing, the real needs put forward by users are no longer simple. So basic services with a single function can hardly meet this demand. Service composition technology emerged under this background. A typical problem in service composition is the binding of each abstract component service in a process to a concrete Web service in the set of candidate services so that the binding result can satisfy a series of local qos constraints and global qos constraints, and the optimal solution is obtained. This typical problem is qos-aware service composition(Zeng et al., 2004). Many achievements have been made in the research on this problem, however, these studies identify that all component services in a service composition process are equally important. However, there are usually some component services that are core component services, that is to say, the quality of their binding results has a decisive and more significant impact on the quality of the overall combination result. The set formed by these core component services is called the core component service set. This is simply called the core service set. In addition, we note that due to the particularity of core component services in the service composition process, there will be some component services that can assist core component services and are closely related to them. This paper refers to these component services closely related to core component services as diffraction associated component services. The set formed by the corresponding diffraction associated component services is called the diffraction associated set. The core service set and the diffraction associated set have an essential influence on the solving of the service composition problem. For example, the core service set plays a leading role in the process of service composition, the performance of the core component services determine the performance of the entire composite service, the core service set and the diffraction associated set have an effect on improving the qos of the final solution result. In the previous related researches, when proposing methods to solve the service composition problem, the influence of core service set and diffraction associated set on problem-solving is often not considered, or even though their impact is considered, they are not considered in-depth and then used in a deeper level. To solve the above problem, this paper proposes a diffractive service composition method. This method adopts the processing method of distinguishing the primary and secondary and splitting the process to solve the problem. Specifically, this method first combines core component services, their corresponding diffraction associated component services and other corresponding component services using the corresponding algorithm for the first stage of combination. Secondly, uses the combined result of the first stage as the reference point, and uses the corresponding algorithm to combine it with the corresponding diffraction associated component services and other corresponding component services for the second stage of combination. After that, and so on until the combined result of the entire process is obtained.