Most of the signal processing for images, audio, etc. is done digitally nowadays. Sampling of a signal is required for digital signal processing. However, sampling introduces error. Shannon’s sampling theorem is the fundamental result which works best for the bandlimited signals. In practice, the signals are rarely bandlimited. Therefore, we used the sampled-data system theory, instead of the Shannon’s theorem, for solving a few signal processing problems. This theory has its root in the hybrid model-matching problem generally discussed in the control literature. The basic requirement in the digital signal processing is to process the signal with minimal (reconstruction) error. Hence, we concentrated on minimizing the reconstruction error mathematically using the sampled-data system theory. In this presentation, I will discuss solutions of the optimal down sampling and causal sampler design problems using the sampled-data system theory. I will also discuss a method for calculation of the reconstruction error.
Dr. Hanumant Singh obtained his PhD from the department of Applied mathematics at University of Twente in Netherlands in 2012, Masters degree in Electrical engineering from Indian Institute of Technology Bombay in 2004 and Bachelors degree in Electronics and communication engineering from University of Rajasthan in 2002. He worked for Sasken communication and Texas instruments during years 2004-2008.