Dynamic QoS Aware Service Composition Framework Based on AHP and Hierarchical Markov Decision Making

Abstract

In recent years, with the continuous development of serviceoriented computing technology, the industry has increasingly high quality requirements for service solving. A dynamic perceptual service composition framework is proposed for solving Web service problems. During the process, the user provides three aspects of information, including user context, user requests, and user preferences. The relevant elements in the problem are divided into multiple logical levels and processed layer by layer. The entire service process is represented through a directed graph, and specific business is represented using a vertex set. A forward step re-planning rule is established to skip business vertices. The experimental results showed that the research method had an error of only 0.071% in solving quality analysis at 20 business vertices, which was lower compared with other methods. In the multi concurrency tolerance test, the research method only experienced 5 crashes during a 500s runtime. In the analysis of packet loss rate during operation, when there was no network fluctuation, the packet loss rate of the research method fluctuated between 0.1% and 0.7%. The research method can provide service discovery results that are more in line with the actual needs and preferences, which can provide better solution results for service composition problems.

Keywords

• Quality of service
• analytical hierarchy process
• Markov decision
• service composition
• directed graph
• re-planning