Local monitoring and maintenance for operational wireless sensor networks

One of the key mechanisms underlying a wireless sensor network (WSN) is to monitor the network itself. Many existing approaches perform centralized analysis and maintenance based on a large amount of status reports collected from the WSN, while others use add-on protocols/modules that not only require extra management cost, but also interrupt the normal operations of targeted WSN applications. Unlike existing work, we propose LoMoM, a new approach of Local Monitoring and Maintenance for a WSN, which combines monitoring operations for the WSN with the operations of a mobile event monitoring, in a manner that is both energy and latency-efficient. LoMoM includes a two-part monitoring architecture: a WSN and a 3G network. Our main interest is in the WSN-part, where we address two important issues: monitoring probable anomalies/faults of the nodes, and link failures. To achieve event monitoring efficiently, LoMoM conducts a prompt local maintenance when such a fault occurs. Fault and event detection status reports are observed by a remote monitoring center using the 3G network. Comprehensive evaluations via both simulations and real-world experiments are conducted to validate the effectiveness of LoMoM.