Data collection from wireless sensor networks using a hybrid mobile agent-based approach

Lack of foresight into the potential future uses of long-term deployments, and limited resources of wireless sensor networks preclude placing all codes on the nodes at run time. Software mobile agents ease this problem by carrying encapsulated code for specific measurement, aggregation and data processing actions. Agent itinerary schemes usually pose agent routing as a variant of the classic Travelling Salesman Problem. This leads to significant inefficiencies because built-up data inflate the agent. We propose a hybrid technique, harnessing the nature of wireless transmission, mobile agent features, and data aggregation, to provide the efficiency of traditional response path optimization techniques in mobile agent itineraries. Simulations demonstrate that the algorithm outperforms traditional mobile agent scheduling approaches in wireless sensor networks.