Adaptation in Pervasive Computing Environments: A Multi-agent approach

Abstract

This paper proposes a multi-agent based context-aware adaptive system that can autonomously adapt the levels of performance of applications/services in event of a significant mismatch between demand and supply of resources in a pervasive computing environment. In event of scarcity of resources, the system autonomously chooses a best-fit adaptation strategy relevant to the context based on user preferences. The primary emphasis here is to continue to provide acceptable levels of a service, avoiding as far as possible, a direct involvement of the user.