In this paper, a novel time-phased directional-sensor network deployment strategy is presented for the mobile-target search problem, e.g., wilderness search and rescue (WiSAR). The proposed strategy uses probabilistic target-motion models combined with a variation of a standard direct search algorithm to plan the optimal locations of directional-sensors which maximize the likelihood of target detection. A linear sensing model is employed as a simplification for directional-sensor network deployment planning, while considering physical constraints, such as on-time sensor deliverability. Extensive statistical simulations validated our method. One such illustrative experiment is included herein to demonstrate the method’s operation. A comparative study was also carried out, whose summary is included in this paper, to highlight the tangible improvement of our approach versus three traditional deployment strategies: a uniform, a random, and a ring-of-fire type deployment, respectively.
Automatic Middleware Deployment Planning On Clusters
