Low-Power Sensor Networks for Real-Time Air Quality Monitoring in Dense Urban Districts

Abstract

Fine-grained air quality monitoring in dense urban districts requires sensor deployments capable of operating for extended periods without frequent battery replacement, while maintaining measurement accuracy comparable to reference-grade equipment. This study presents a low-power sensor network architecture combining duty-cycled particulate and gas sensors with an adaptive sampling protocol that adjusts transmission frequency based on detected pollutant variability rather than fixed intervals. Field deployment across several urban districts with contrasting traffic density allowed comparison of network energy consumption and data completeness against a fixed-interval baseline configuration. Results show that adaptive sampling substantially extends node operational lifetime while preserving detection of short-duration pollution spikes associated with traffic congestion events, which fixed-interval sampling at comparable average energy budgets tended to miss. Periodic co-location with reference-grade monitoring stations enabled calibration drift correction, revealing that low-cost sensor accuracy degraded gradually over the deployment period in a pattern predictable enough to support scheduled recalibration rather than continuous reference comparison. These findings support the feasibility of dense, long-duration low-cost air quality sensor networks for urban districts lacking comprehensive reference-grade monitoring infrastructure, with practical guidance for balancing energy budget against temporal resolution requirements.