Abstract:

Aiming at the problem that the importance of signal contour and the effectivity of signal are ignored in analog to information conversion (AIC). Based on contour pre-extraction, a new AIC and reconstruction scheme is presented within the compressed sensing theoretical framework. The contour of the input signal is pre-extracted by a low speed analog to digital converter （ADC） and then the details are compressed. To recover the real-time streaming signal exactly, an adaptive piecewise orthogonal matching pursuit algorithm is proposed. The theoretical analysis proves that the contour pre-extraction AIC is effective and reliable, and the validity of theoretical analysis results is demonstrated through simulation. It is shown that the performance of signals recovery can be greatly improved by the contour preextraction AIC with a bit of signal contour information.