Enhancing Multi-hop Sensor Calibration with Uncertainty Estimates

Low-cost sensors, installed on mobile vehicles, provide a cost-effective way for fine-grained urban air pollution monitoring. However, frequent calibration is crucial for low-cost sensors to consistently deliver accurate measurements. Multi-hop calibration is a common practice to calibrate mobile sensor deployments, but is prone to severe error accumulation over hops. Prior research mitigates error accumulation by designing special calibration models, which only apply to linear models. In this paper, we propose an orthogonal approach by selecting reliable measurements for calibration at each hop. We analyze the impact of different data-induced uncertainties on calibration errors and devise a scheme to estimate these uncertainties of the calibrated outputs. We further propose an uncertainty-based metric for data filtering at each hop. We evaluate the effectiveness of our method in a real-world ozone sensor deployment. Experimental results show that our method works with both linear and non-linear calibration models and reduces calibration errors in multi-hop setups by up to 25% compared with existing techniques.