Energy Efficient Wake Up Radio Polling Based on Value of Information

The Wake-Up Radio (WUR) system enables battery-powered sensors to stay in low-power sleep mode while listening for a wake-up signal (WUS). Upon receiving a WUS, the sensor activates its primary communication radio (PCR) to transmit data, reducing idle listening and retransmissions. However, polling data inefficiently can waste energy and storage, especially when updates provide little meaningful information. Due to the sink’s limited knowledge of node-side processes, accurately selecting nodes that provide valuable updates is challenging. We propose a Whittle Index AoII (WAoII) policy with state estimation techniques, allowing optimal node selection without assuming transition probabilities. Experimental results show WAoII with state estimation achieves near-optimal performance, reducing packet transmissions by up to 70% while enabling accurate data pattern reconstruction an aspect overlooked by other methods. These findings highlight WAoII with state estimation as an effective approach for polling data from wireless sensor nodes (WSNs) with WUR, maximising the value of information (VoI) while minimising energy and storage demands.