Performance evaluation of cluster-based target tracking protocols for wireless sensor networks

Abstract

Target tracking is an important application type for wireless sensor networks (WSN). Recently, various approaches are proposed to maintain the accurate tracking of the targets as well as low energy consumption. Clustering is a fundamental technique to manage the scarce network resources. The message complexity of an application can be significantly decreased when it is redesigned on top of a clustered network. Clustering has provided an efficient infrastructure in many existing studies. The clusters can be constructed before the target enters the region which is called the static method or clusters are created by using received signal strength (RSS) from target which is called the dynamic method. In this paper we provide simulations of static and dynamic clustering algorithms against various mobility models and target speeds. The mobility models that we applied are random waypoint model, random direct model and Gauss Markov model. We provide metrics to measure the tracking performance of both approaches. We show that the dynamic clustering is favorable in terms of tracking accuracy whereas the energy consumption of static clustering is significantly smaller. We also show that the target moving with Gauss Markov model can be tracked more accurately than the other models.