A distributed preventive maintenance strategy for wireless sensor networks

In this paper, we have proposed two distributed preventive maintenance strategies dedicated for large-scale wireless sensor networks (WSNs) using a reduced number of mobile robots : (i) Fixed Distributed Preventive Maintenance Strategy denoted by F-DPMS, (ii) Adaptive Distributed Preventive Maintenance Strategy denoted by A-DPMS.

F-DPMS and A-DPMS are based on a simple analytical model to estimate the expected failures. To share the sensors repair load among the available robots, F-DPMS subdivides the WSN sensing area in equal surface subareas. Each maintainer robot is assigned to a single one area to deal with its anticipated failures. In contrary, A-DPMS uses an adaptive area partitioning technique depending on the distribution of failures over the network map, in order to balance the repair load sharing. Simulation has shown that both F-DPMS and A-DPMS provide a null network dysfunction time with a sufficient number of robots. Obtained results show also that A-DPMS requests a minimal number of robots lower than F-DPMS. But, it produces a heavier signaling cost. To remedy this problem, we have proposed a new version of A-DPMS denoted by A-DPMS*. The novel maintenance strategy reduces not only the introduced signaling messages, but also the traveled distance achieved by the maintainer robots to replace the expected failures.