Differential on-line sensor calibration [light intensity sensor example]

We have developed an on-line calibration scheme that employs a single source as the external stimulus that creates differential sensor readings used for calibration. The technique utilizes the maximal likelihood principle and a nonlinear system optimization solver to derive the calibration function of arbitrary complexity and accuracy. The approach is intrinsically localized and we present two variants: i) one where only two neighboring sensors have to communicate in order to conduct calibration; ii) one where a provably minimum amount of communication is achieved. We evaluate the techniques using traces from light sensors recorded by deployed sensors, and statistical evaluations are conducted in order to obtain the interval of confidence to support all the results.