一种基于QoS的Web服务组合算法

摘要

针对基于响应时间的Web服务组合，提出一种迭代加深的A*算法（IDA*算法）对服务节点进行搜索筛选。在利用节点数量的要求设计出服务节点限制判断条件的基础上，将该 IDA*算法进行具体实现。实验结果表明，与图规划搜索算法相比，该IDA*算法在所需目标节点数较少的范围内比图规划搜索效率高。还进一步将IDA*改进为基于top-k的算法，并进行分析比较。%For Web service composition based on response time,we propose an iterative deepening A*algorithm (an IDA*algorithm)to search and screen the service nodes.We use the requirement of service nodes number to design the determination condition of service nodes limitation,on that basis we give the specific implementation of the IDA* algorithm.Experimental results demonstrate that compared with graph-planning search algorithm,the IDA* algorithm is more efficient than it in the range of just asking for a small number of goal notes.In this paper,we further expand the IDA* algorithm to the top-k based algorithm,and make the comparisons and analyses as well.