A skyline-based efficient web service selection method supporting frequent requests
Yue Wang, You Song, Mingyang Liang · 2016
With the development of web technology, a sharp increment in the number of services brings in efficiency problem to service selection. Furthermore, in dynamic service selection, services have to be selected as soon as requests arrive. When requests to a service are frequent, the service is prone to failure. This paper proposes a skyline-based web service selection method to deal with efficiency problem as well as solve the frequent requests problem. The method covers a search algorithm and a service scheduling strategy. To improve efficiency of service selection, the search algorithm based on Kd-tree is designed to determine skyline and then narrow search space. To solve the frequent requests problem, lottery scheduling strategy is employed to schedule services. Every service will get a certain amount of tickets in the lottery scheduling strategy, and service with more tickets will be distributed with more requests. Experiments validate the performance of the method.