Computing an Optimal Path with the Minimum Number of Distinct Sensors

Wireless sensors can be deployed in boundary regions to replace human in intruder detection. Given a deployment of such a wireless sensor network, we present a method to compute a path crossing the minimum number of distinct sensors from a source to a destination. Then, we propose the selfish path problem, in which the intruder is always getting more closer to the destination, and we design a greedy algorithm for this problem. We implement these algorithms and compare the number of sensors crossed by a random path, a line segment path, a selfish path, and two paths generated by Dijkstra's algorithm and an adapted Bellman-Ford algorithm. Some important empirical results are obtained.