Research On Energy-Efficient Target Tracking In WSNs

WSNs(wireless sensor networks) is a sort of new network which consists of large number of sensors, it is another hotspot since the Internet technology and is used to collect information from where the WSNs covers. With the development of sensor, computer, micro-mechanical electronic and wireless communication techniques, the WSNs is more and more popular in daily life. Generally,the WSNs can be randomly deployed to form a large scale of coverage, and the network is self-organized in which the sensors can adaptively schedule their activation according to the environment. These advantages of the WSNs make it popular in vehicle-tracking, environment -inspection, military-use and business purpose, etc.However, the WSNs has the inherent limitations of constrained- energy, extreme large number and density of sensor nodes, low reliability of wireless communication which make it quite different from common networks. All these disadvantages call for efficient usage of the constrained energy of the network.With maximizing network's lifetime as our primary purpose, we investigated the target-tracking application and proposed the SAT(Speed Adaptive Tracking Algorithm)algorithm aiming at single target tracking and the COTA(Continuous Object Tracking Algorithm)algorithm which aims at continuous object tracking。Most prediction-based tracking algorithms in single target tracking are cluster-based which lead to large computation in sensor nodes and massive wireless communication which will greatly decrease the network's lifetime by consuming energy overly. To get rid of these shortages, the SAT algorithm can avoid complicated computation and network clustering by adaptively scheduling the nodes'activation according to the target's current speed. Simulation shows that the SAT algorithm is excellent with its tracking efficiency and it can greatly decrease the unnecessary energy waste in the networks.In succession, we proposed COTA algorithm to track continuous objects such as wild fire, birds migrating and oil leaking which differ from single or multiple targets, as they probably occupy 2D or 3D space continuously. In COTA, we use a mechanism similar with what we use in the SAT algorithm: avoid unnecessary communication and computation by analyzing received packages, from which the nodes can predict the object's motion in next time slot. Compared with other continuous object tracking protocols, COTA can prolong the networks'lifetime by sleeping more outlying sensor nodes. Simulation shows COTA can still track object steadily while object's shape, location and area change greatly.In the last section of paper, we developed the Fire Watcher System based on the WSNs,in which we embedded the MAC protocol used in our target tracking research, and this protocol contains a sleep-polling mechanism which can prolong our system's lifetime greatly.