A Finite Queue Model Analysis of PMRC-based Wireless Sensor Networks

In our previous work, a highly scalable and faulttolerant network architecture, the Progressive Multi-hop Rotational Clustered (PMRC) structure, is proposed for constructing large-scale wireless sensor networks. Further, the overlapped scheme is proposed to solve the bottleneck problem in PMRC-based sensor networks. As buffer space is often scarce in sensor nodes, in this paper, we focus on studying the queuing performance of cluster heads in PMRC-based sensor networks. We develop a finite queuing model to analyze the queuing performance of cluster heads for both non-overlapped and overlapped PMRC-based sensor network. The average queue length and average queue delay of cluster head in different layers are derived. To validate the analysis results, simulations have been conducted with different loads for both nonoverlapped and overlapped PMRC-based sensor networks. Simulation results match with the analysis results in general and confirm the advantage of selecting two cluster heads over selecting single cluster head in terms of the improved queuing performance.