Agent service matchmaking algorithm for autonomic element with semantic and QoS constraints

摘要

Agent service description and matchmaking problem for autonomic element has been taken as one of the most important issue in the field of autonomic computing based on agent and multi-agent system. Considering the semantic and QoS for capability description of agent service are two important issues during matchmaking, the factors of semantic and QoS during matchmaking are considered together, and an agent service description model named ASDM_SQ is proposed. On basis of this model, a matchmaking algorithm with semantic and QoS constraints named ASMA_JQ, is presented to find the agent service satisfied both the given semantic similarity threshold and optimal QoS performance. The proposed algorithm lies over two fundamental processes: semantic similarity matchmaking and QoS matchmaking. During QoS matchmaking, evaluation mechanism of confidence of individual QoS attributes, i.e. fidelity factor is introduced to overcome drawbacks such as subjectiveness and unfairness and improve the self-configuration capability for agent element. Simulation experiments demonstrate the effective and correction of our algorithm for agent service matchmaking, and perform better QoS performance than other existing algorithm, which can further show that our algorithm has better compromise between attribute quality and users' evaluation when selecting agent service.