Sensor Deployment for Collaborative Target Detection in the Presence of Obstacles

Sensor deployment is an important issue for surveillance networks especially when obstacles are present in the monitored region. An intruder can hide beyond obstacles or stay at a low-covered location in order to reduce the probability of being detected. A good sensor deployment should maximize the worst-case detection probability over all possible locations of the intruder. Many heuristics have been proposed in the literature for sensor placement to achieve better detection performance. However, none of them deploy sensors based on the optimal distribution of sensors tailored for the given terrain. In this paper, a sensor deployment approach is proposed using the optimal sensor distribution as a reference. The optimal sensor distribution for the given terrain is first calculated and, then, a clustering-based approach is developed to guide sensors to appropriate locations. The effectiveness of the proposed approach is shown by simulations for regions with and without obstacles. The final deployments show that some sensors' final locations are very close to the obstacles. This is in contrast to the conventional assumption used in many previous studies.