Target Tracking in Wireless Sensor Network
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
quotTarget tracking is significant application of Wireless Sensor Network (WSN), necessitates systematized groups of sensor nodes which predict and track the moving target or measures precise distance from the sensor nodes to the target and localize it. Due to limited hardware and processing capabilities of WSN, it is difficult to develop target tracking mechanism which gives precise, fast and reliable localization along with good tracking efficacy.
newline
newlineUncertainties in target motion and limited sensing regions of sensor causes single-sensor-based collaborative tracking suffers from low tracking accuracy and lack of tracking reliability when a target can t be detected by sensor nodes. Actuating multiple sensors can achieve better tracking performance but with high energy consumption. Adaptive scheduling of multiple groups of sensors called clusters is used to get energy efficient tracking while maintaining the tracking accuracy. This is done through designating one of the sensors as the cluster head for estimation update and for scheduling of sensors according to energy measure function called Cluster Head Energy Measure (CHEM). Cluster heads may execute Particle Swarm Optimization (PSO) which maximize and minimize the network coverage and the energy metric respectively and helps to achieve energy efficient tracking.
newline
newlineConventional target tracking models use mathematical algorithm referred to as a state estimator or prediction filter which estimates future state of a moving target depending on its past records. Kalman Filter (KF) being best linear minimum variance sequential state estimator. KF is an optimal filter in mean square error sense, if the noise covariances are known to the filter prior to its operation. If these values are taken wrong, then KF may diverge. To avoid so, PSO is used to update covariance iteratively. For non linear model, Extended Kalman Filter (EKF) could be used.
newline
newlineParticle filter (PF) is considered to be the most accepted filtering algorithm in various tracking and localization problems cha