|
Abstract:
|
We propose a sequential Bayesian approach to distributed source localization in wireless sensor networks using the Received Signal Strength (RSS) measurements that follow an energy-decay model. The sequential approach to the problem allows the processed measurements to be communicated across the sensors in a distributed fashion that offers flexibility in managing the communication cost. We consider several probabilistic fusion schemes, namely, the Independent Likelihood Pool, the Independent Opinion Pool and the Linear Opinion Pool. We also consider some numerical approximations that can reduce the communication cost. We evaluate the performance of the proposed methods and compare them with existing distributed algorithms such as incremental non-linear least-squares approach and show that our methods outperform them in terms of the MSE and communication cost.
|
