Poisson Versus Gaussian Distribution for Object Tracking in Wireless Sensor Networks

Object tracking is one of the fundamental applications in Wireless Sensor Networks (WSNs). To detect and track the appearance and movement of malicious object(s), a number of sensors are usually deployed randomly in the area of interest specially for hostile application scenarios. Following random deployment strategy, the resulting WSNs conform to Poisson or Gaussian distribution, depending on specific deployment approach. It is of significant importance to investigate and compare the performance of Poisson and Gaussian distributed WSNs for object tracking, since sensor deployment plays an important role in the QoS of WSNs. In view of this, this paper firstly captures the problem and stochastically evaluates and compares the QoS of Poisson and Gaussian distributed WSNs for object tracking. Two detection models are employed: single-sensing detection and multiple-sensing detection in the analysis. Effects of different network parameters on the detection capability of Poisson and Gaussian distributed WSNs are evaluated for comparisons. This exploration leads to a spectrum for selecting appropriate deployment strategy in designing application-specific WSNs such as object tracking.