TTS-Coded Genetic Algorithm for QoS-driven web service selection

摘要

To maximize user satisfaction during composition of web services, a novel genetic algorithm based on a special tree traversal sequence (TTS) coding scheme is presented in this paper. Utilizing the tree traversal information toted by individuals of genetic population and the composite logical information stored in tree composite pattern (TCP), TTS coding scheme can support all the 4 basic models of service composition flows in a much simpler and more effective way. Additionally, an adaptive crossover and mutation strategy is adopted to improve the algorithm's search capability and constringency speed. Experiments show that the adaptive TTS-coded genetic algorithm (GA) can get more excellent composite service plan than matrix-coded and tree-coded GA.