An itinerary for data fusion in distributed sensor network

Using mobile agent for data fusion in distributed sensor network brings new problems, the serial itinerary of mobile agent travel makes long task duration, mobile agent transfer introduces new overhead, if a mission need collect data from all nodes many times mobile agent must be dispatched many times, which brings much data transfer overhead. To overcome these shortcomings, A closest first (CF) algorithm is presented to calculate a binary full spanning tree (BST) itinerary for mobile agent, full spanning tree ensuring shortest path. And the function of mobile agent is changed, when first reaching a node mobile agent give a copy of its code (called Resident Agent) into the node, resident agent can stay in the node for many times data collections and data fusion. Simulation shows that resident agent based data fusion can both reduce energy consumption and task duration time.