A Self-Evolutionary Method for Service Emergence in Pervasive Computing Environments

The paper proposes a self-evolutionary approach of service emergence inspired by the characteristics of biological neuroendocrine and immune system. In pervasive computing environments, services are represented by autonomous bio-entities (mobile agents with biological operation), and bio-entities establish relationship network based on affinity to provide service emergence. Affinity adjustment depends on three factors, namely, the matching strength of bio-entities, the service quality score, and the trust. The method completes a series of work from composition to management autonomously. The simulation results show that the approach can form relationship clusters and significantly improve the service emergence performance. It adapts well to the changes of dynamic environments.