Routing mobile agent to local regions for data fusion in distributed sensor networks

The problem of computing a route for a mobile agent to local regions or specified nodes in a distributed sensor network is considered, both in a single region or multi-region cases. With all specified nodes visited, this paper formulates the route computation problem in terms of minimizing an objective function, which is directly proportional to the energy consumption and path loss. Further, we propose heuristic algorithms based on local closest first methods, shortest path algorithm and algorithms for Steiner tree problem, and an improved ant colony algorithm to compute a suboptimal solution. The proposed ant colony algorithm enhances convergence speed while ensuring solution quality by adjusting pheromone decay parameter rho and importance of exploitation versus exploration adaptively. Simulation results are presented for comparison between the two kinds of algorithms.