Processor Optimization of an Energy-Efficient NDIR CO2 Wireless Sensor Node

Abstract

CO2 gas sensors are rapidly growing in importance since they can be easily deployed to assess air quality inside buildings, improving people's life. According to medical studies, we know that exposure to high levels of carbon dioxide on a daily basis poses a risk to people's health with adverse effects at different levels, such as on the neurological, cardiovascular, and respiratory systems. Moreover, the COVID-19 pandemic has shown how important it is to monitor air quality and ensure good ventilation to prevent infection through airborne transmission paths. Some studies estimate the likelihood of indoor airborne infection transmission based on continuous CO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> measurements. This work presents the design of a 32-bit microprocessor based on the RISC-V ISA architecture intended for energy-efficient signal processing in wireless sensor nodes. In particular, the processor is optimized for its application in non-dispersive infrared (NDIR) CO2 sensors. We use asynchronous demodulation to extract the information from the sensor. Therefore, we generate digital quadrature signals for demodulation, requiring the computation of trigonometric functions by the RISC-V processor. We try different strategies to optimize the processor design and the demodulation process to minimize energy consumption while measuring CO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> levels.