Exploiting dynamic weaving for self-managed agents in the IoT

Abstract

Agents are a good option to develop Internet of Things (IoT) systems with self-management capacities. Current agent technology offers the necessary means to manage many of the requirements of self-management satisfactorily, however agent approaches for lightweight devices do not provide support for self-configuring agents' internal functionality adequately. This paper describes part of the Self-StarMAS approach for self-managing the IoT, which uses aspect-oriented mechanisms to model self-managing tasks. Aspect-orientation facilitates the adaptation and reconfiguration of the Self-StarMAS agents' internal architecture. This paper focuses on describing the internal composition of aspects, which is performed by an internal weaver at runtime. © 2012 Springer-Verlag.