Wireless sensor networks (WSNs) are an appealing choice for applications such as environmental monitoring due to the capability to deploy massive amounts of cheaper sensor nodes and manage them in a distributed fashion. An important problem in this field is how to deal with simultaneous failures of sensor nodes. It is not difficult to imagine a natural disaster or critical condition rendering a potentially large set of nodes within a vicinity inoperable. How does one recover from that?

In this paper, the authors take a specific use case--precision agriculture--and investigate a solution. Such a network consists of cheaper static sensor nodes (SNs) and relay nodes (RNs) with unlimited power supply. The RNs are in charge of a polygon region of SNs to collect, aggregate, and forward to sinks or final data collection centers. The authors view the recovery of simultaneous failures as a multi-objective optimization problem (MOP) and use a heuristics algorithm to redeploy the RNs to reconstruct the polygons. The algorithm is called simultaneous failure recovery based on relay node relocation (SFR-RNR). The primary idea is to move some RNs toward the center of the failure region and form “larger” polygons to compensate and resume connectivity. For this to happen, there is of course a process of self-volunteering and negotiation and back-off among RNs. In order to show its effectiveness, various aspects of the proposed SFR-RNR are compared with an existing algorithm (DORMS) from prior literature, including the number of relocated RNs, the percentage of coverage, total travel distance of nodes after redeployment, the number of redeployed sensors, and the effects of the number of failed RNs and the percentage of damaged area.

However, the investigated use case does not feel general enough to me--the network is deployed in a very organized fashion and is equipped with relay nodes with unlimited power supply. For applications outside precision agriculture, the deployment pattern may not benefit as much from this approach.