On real-time performance evaluation of volcano-monitoring systems with wireless sensor networks

A volcanic eruption early warning has to be launched with effectiveness and within the shortest time possible, which imposes the requirement of using real-time (RT) systems. In this setting, volcano-monitoring systems using wireless sensor networks (WSNs) may play a key role. Previous work did not report detailed-enough performance evaluation, in order to identify their main constraints as RT systems, either in simulation tools or in testbed scenarios. The aim of this paper was to identify the optimum number of sensors to be deployed a posteriori, based on simulation results considering throughput, packet loss, and end-to-end delay, as metrics to satisfy the RT requirements. We corroborated the simulation results obtained by a testbed deployment within a controlled environment. We determined that optimal scenario for volcano monitoring is a random topology, and the results show that 12 nodes should be deployed as maximum to satisfy the RT constraints. To test the system in a real scenario, 10 sensors were deployed in a strategic area at Cotopaxi Volcano, and information was collected during three days of continuous monitoring. This information was sent to a remote surveillance laboratory located 45 km away from the station placed at the volcano using WiFi-based long-distance technology. Our study shows that the coordinator node is the main bottleneck in the real application scenario, given that its processing rate provokes an excessive time delay near to 3 s, which has to be solved to satisfy the RT requirements. We conclude that a comprehensive study, including simulation, testbed, and in-situ deployment, provides valuable information for the specifications to be accounted in permanent WSN RT volcano monitoring.