Nonlinear modulation with low-power sensor networks using undersampling

Nonlinear modulation is a promising technique for ultrasonic non-destructive damage identification. A wireless sensor network ideally suited to monitor large structures using nonlinear in cost-efficient manner. However, existing approaches rely on high sampling rates and resource-demanding computations that are not feasible low-cost low-power devices. We present new indicator uses the short-time Fourier transform derive amplitude phase with less computational effort memory usage. Evaluation of proposed method real experiment data exhibits performance reliability similar conventionally used index. Undersampling demonstrated, which reduces demand test scenario by more than 100 times, required energy processing four times. The loss accuracy introduced undersampling shown be negligible.