TT-Mote: An Architecture towards Perpetual Wireless Sensor Networks

Most sensor nodes are battery-powered, and often need periodic replacement, limiting their usage in remote places. Various energy harvesting sensor nodes and different power optimization algorithms, like deep sleep methods, are in practice to address this problem. Even in deep sleep mode, most sensor nodes consume significant power as this sleep duration is usually longer than the standard operating time. This paper presents a novel method that harvests solar energy, uses BQ25570 for energy management, a supercapacitor to store energy, and uses LoRa for communication. The proposed architecture employs programmable timers to shut down the sensor node, reducing power consumption and eliminating energy in deep sleep mode. The timer-triggered sensor node's activity was monitored for 15 days, transmitting data every 10 minutes. It is observed that even in low-light conditions, the sensor node may operate for more than seven days. Considering a periodic wake-up at every 10 seconds and sleep duration of 3 seconds, it was found through experimentation that a timer-triggered sensor node conserves 595 milliwatts of energy every day compared to a sensor operating in a deep sleep mode.