Cloud Service Composition and Optimization Selection Based on Hybrid Service Composition Algorithm

Abstract

In the cloud environment, there are many cloud services with the same function, but there are differences in performance and quality of service (QoS) attributes, so solving service composition and optimization selection (SCOS) in the cloud environment is a complex problem. This study proposes a new method to solve this problem, namely hybrid service composition algorithm, which uses the hunting mechanism of standard ant lion optimization algorithm to find the best service composition path. In this method, on the premise of adapting to the cloud environment and quality of service, crossover and mutation operators are introduced to carry out genetic operation on the population and generate offspring individuals, thus increasing the population size and finding the optimal solution more quickly. In each iteration, the best composite cloud service is selected, and other cloud services are combined according to it. By adjusting genetic factors and iteration times, the population can be better improved and the local optimization of the service composition fitness value can be avoided. In this study, we test and prove the effectiveness of the algorithm for SCOS problems of different sizes. Experimental results show that, compared with GWO algorithm and Ant Lion algorithm, the proposed algorithm has better performance and can find better solutions, and improves accuracy and stability, especially in dealing with large-scale SCOS problems.