In this paper we study the problem of tracking an object moving randomly through a network of wireless sensors in the presence of clutter. Our objective is to devise strategies for scheduling the sensors to optimize the tradeoff between tracking performance and energy consumption. The presence of random interference introduces uncertainty into the origin of the measurements. Data association techniques are thus required to associate each measurement with the target or discard it as arising from clutter (False alarms). We cast the scheduling problem as a Partially Observable Markov Decision Process (POMDP), where the control actions correspond to the set of sensors to activate at each time step. Exact solutions are generally intractable even for the simplest models due to the dimensionality of the information and action spaces. Hence, we develop an approximate sensor scheduler that optimizes a point-based value function over a set of reachable beliefs. Point-based updates are driven by a non-linear filter that combines the validated measurements through proper association probabilities. Our approach efficiently combines Probabilistic Data Association techniques for belief update with Point-Based Value Iteration for designing scheduling policies. The generated scheduling policies, albeit suboptimal, provide good energy-tracking tradeoffs.
展开▼