Maximum lifetime suspect monitoring on the street with battery-powered camera sensors

Abstract

A camera sensor network is a sensor network of a group of camera sensors and is being deployed for various surveillance and monitoring applications. In this paper, we propose a new surveillance model for camera sensor network, namely half-view model, which requires a camera sensor network to capture the face image of any object if it moves forward to pass over an area of interest. Based on this new surveillance model, we introduce a new sleep-wakeup scheduling problem in camera sensor network, namely the maximum lifetime half-view barrier-coverage (MaxL-HV-BC) problem, whose goal is to find an on-off schedule of battery-operated camera sensors such that the continuous time duration providing half-view barrier-coverage over an area of interest is maximized. We develop a strategy to check if a region is half-view covered by a given set of camera sensors, and use this strategy to design two new heuristic algorithms for MaxL-HV-BC. We also conduct simulations to compare the average performance of the proposed algorithms with a trivial solution as well as the theoretical upper bound.