Simulated study of the influence of node density on the performance of wireless sensor networks

This paper investigates the impact of local and global node density in cluster-based structured wireless sensor networks (WSNs). The local density represents sensor node density (SND) in a cluster whereas, global node density relates to head node density (HND) in the entire WSN. The literature rarely addresses the impact of density on WSNs performance as the focus is typically on protocols, routing, scheduling, clustering and network longevity. Often, the density of nodes is assumed heuristically, but not based on empirical experiments. In this work, we address this issue by measuring the impact of node density on four performance metrics: isolated sensor nodes, isolated head nodes, network detection effectiveness and network tracking accuracy. Using an in-house simulator, a total of 5,200 experiments were conducted and performance-metrics were collected and analysed. The results revealed interesting relationships among the studied variables and identified best performing node densities locally and globally.