Learning Algorithms for Complete Targets Coverage in RF-Energy Harvesting Networks
IEEE Transactions on Vehicular Technology
Internet of Things (IoTs) networks are responsible for monitoring an environment or targets such as vehicles. A key issue is determining the active time of a set of sensor nodes, so called set cover, that monitors all targets. This requires battery level knowledge at sensor nodes as an incorrect active time may cause energy outage, leading to uncovered target(s). However, in practice, it is impractical to obtain this information, especially in large-scale networks. To this end, we present a number of approaches to construct set covers. We first propose a Two-Phase Algorithm (TPA) that requires sensor nodes to first determine their probability of being active in each time slot. This information is then used by the HAP to construct set covers. We then introduce learning approaches based on Gibbs and Thompson sampling. The Gibbs sampling based algorithm or GB allows a sink/gateway to learn the best set cover to use over time. Similarly, our Thompson sampling solutions, namely TS-Random and TS-CB, construct set covers iteratively based on the success probability of sensor nodes in monitoring targets. The numerical results show that TS-CB converges to the optimal solution. GB performs better than TS-CB initially but has similar performance to TS-CB in the long term.
Open Access Status
This publication is not available as open access