Smart Decision Making Policy for Faster Harvesting From Ambient RF Sources in Wireless Sensor Nodes
Résumé
For wireless sensor network nodes (WSNNs), which
are usually distributed over wide area and transmit intermittently,
radio frequency energy harvesting (RFEH) from ambient
sources such as base stations, TV towers, access points etc. seem
to be practical and promising approach. Such RFEH enabled
WSNNs need intelligence to choose optimal frequency band(s) for RFEH from wideband spectrum. Furthermore, for minimizing energy consumption, number of frequency band switching (FBS) should be as minimum as possible. To meet these requirements,
a new multi-stage decision making policy (DMP) for RFEH
enabled WSNNs has been proposed in this paper. The proposed DMP uses Bayes-UCB algorithm to identify the subbands with high RFEH potential. Another novelty of the proposed DMP is that RFEH duration is made tunable based on the RF potential
of the chosen subband(s) leading to significant improvement in harvested energy and number of FBS. Simulation results show
that the proposed DMP offer 10%–18% improvement in total harvested RF energy. Furthermore, the number of FBS in the
proposed DMP are 25%–50% of that in existing DMPs making the proposed DMP a preferred choice for resource-constrained
WSNNs.