A multi-agent hybrid protocol for data fusion and data aggregation in non-deterministic wireless sensor networks

Abstract

Data fusion deals with collaborative in-network processing and gathers relatively accurate information about the events in the environment. Conventional data fusion algorithms when assisted with mobile agent technology shifts computationally intensive tasks to these intelligent units thereby increasing the lifetime of the network. There exist mobile agent based event driven protocols for accumulating and forwarding the information to the sink(base station) eg: Tree-based Itinerary Design [4]. Usually, such protocols either deploy value-based fusion or decision-based fusion but very few are using both at the same time. Moreover, use of multi-agent systems(MAS) in such protocols is still in its infancy. The focus of this work is thus to propose a multi-agent hybrid protocol exploiting the benefits of both value and decision fusion by performing aggregation at the source level in a clustered WSN.