Modeling and Evaluating the Reliability of Cluster-Based Wireless Sensor Networks

Abstract

The reliability of wireless sensor networks (WSNs) depends on both sensing coverage and reliable transmission of collected data provided by the target cluster of sensors in the proximity of the phenomenon to the observer. In this paper, we establish a probabilistic fundamental quantitative notion for performance-critical applications on reliable information transfer and propose a new analytical model to calculate the coverage-oriented reliability of cluster-based WSNs. Here, the packet loss probability is modeled as a function of two main parameters. The first is the probability of link failure when the node's buffer at the end point of the link is full (discarding the packet) or when the link encounters some errors (corrupting the packet). The second is the probability of path failure when at least the energy of one node is depleted on the path. Most of network properties are considered as random variables and a queuing-theory based model is derived. Simulation results are used to validate the proposed model. The simulation results agree well with the model.