No abstract

An improved random equivalent sampling (RES) approach based on compressed sensing (CS) for repetitive ultrasonic signal detection is presented. The proposed system considers recovering ultrasonic signal with high equivalent sampling frequency from samples captured using analogue-to-digital converter (ADC) clocked at a rate much lower than Nyquist rate. A basis function is constructed to realise the ultrasonic signal sparse representation, which paves the way for applying CS theory to ultrasonic signal sub-Nyquist sampling. A sampling architecture is developed, which is applicable for ultrasonic compressive detection. The CS measurement matrix is constructed in the context of RES. Experimental results indicate that, with this sampling system, an ultrasonic signal sampled at ultra-low rate, but still can be recovered with overwhelming probability.