The Improved-Iterative Support Detection Algorithm for Pulse-Position-Modulation ADC Architecture

Compressed sensing permits sub-Nyquist sampling, breaking the bottleneck of traditional Nyquist theory. The random pulse-position-modulation analog-to-digital converter (PPM ADC) is a sub-sampling structure, employing compressed sensing theory. To reconstruct signal in PPM ADC, period random sampling reconstruction (PRSreco) algorithm is firstly used, but it requires signal's prior sparsity information. Iterative support detection (ISD) algorithm recovers signal with no need of sparsity, thus we apply it to PPM ADC. To obtain a more accurate recovery especially when sub-sampling ratios are low, we propose an improved-ISD algorithm in this paper. Based on ISD, we add a SupportCorrection module and a denoising module. The former is implemented to choose more appropriate frequency support and the latter is carried out to eliminate the burr of frequency coefficients. Simulation results demonstrate that improved-ISD increases the recovery robustness at low sub-sampling ratios and augments mean output SNR.