Adaptive edge detection with distributed behaviour-based agents in WSNs

Abstract

Wireless sensor networks have emerged as fundamentally efficacious tools for monitoring widespread environmental phenomena. To accurately sense and duly track these time-varying phenomena, it may be necessary for each node to locally determine whether it lies at region edge of these phenomena. However, existing researches show that edge detecting techniques are non-trivial to design and implement in arbitrarily deployed sensor networks. In this paper, we propose a novel computational method of edge detecting for wireless sensor networks which utilises a collection of distributed autonomous agents. The goal of the agent-based edge detecting method is to efficiently locate and label various homogeneous regions of the monitored field. Therefore, these autonomous agents will reside in the individual sensor node and execute a series of reactive behavioural responses. We evaluate the performance of the adaptive edge detecting method and conclude that the proposed method readily outperforms its existing counterparts.