Synchronous transmissions + channel sampling = energy efficient event-triggered wireless sensing systems

Abstract

Wireless Sensor Networks have seen consistent improvements of their end-to-end latency, reliability and power efficiency; thus making possible novel applications of event-triggered systems. To date, the best performance has been consistently demonstrated by protocols that rely on floods of synchronous transmissions (e.g., Glossy, Back-to-Back Robust Flooding, Crystal). However, these protocols use the flooding primitive both for network wake-up and event notification, periodically probing the channel in order to receive a flood in case an event has been detected. As floods are energy expensive, their extensive use for probing the network results in high power consumption and spectrum usage, being the dominant source of energy use in low event frequency scenarios. Moreover, the frequent flooding may hamper the network co-existence, thus representing a key obstacle for event-triggered applications. Our paper proposes Synchronized Channel Sampling (SCS), a reliable wake-up primitive that is capable of reducing the energy consumption of protocols by replacing the floods-based probing with a synchronous channel sampling wake-up mechanism. The testbed experiments performed show that SCS brings power reductions of 33.3%–40% to the state-of-the art protocol B2B (winner of the dependability competition EWSN'17), while maintaining equivalent reliability performance and reducing the spectrum usage.