RAND PPM : A LOW POWER COMPRESSIVE SAMPLING ANALOG TO DIGITAL CONVERTER
Presented by: Praveen Kumar Yenduri, Author(s): Praveen Yenduri, Anna Gilbert, Michael Flynn, Shahrzad Naraghi, University of Michigan, United States
Analog-to-digital converters that digitize time information are known for their low power consumption. Coupling them with sampling techniques that take advantage of the signal compressibility leads to a further efficient data conversion. In this direction, we propose a new analog-to-digital converter, rand PPM, that employs compressive sampling techniques to efficiently sample at sub-Nyquist rates. The PPM (pulse-position-modulation) architecture uses a periodic ramp signal as reference and compares it with the analog input signal to eventually measure the ramp crossing times which determine the signal amplitudes. By appropriately introducing randomness into the reference ramp signal the PPM ADC architecture can be modified into a compressive sampling analog-to-digital converter. The sub-sampled signal is reconstructed using the developed algorithms tailored for practical hardware implementation. We have developed a theoretical analysis of the hardware system and reconstruction algorithm along with a suite of numerical experiments that support the theory.