Intentional Aliasing Method to Improve Sub-Nyquist Sampling System

TL;DR

This paper introduces an innovative aliasing technique in sub-Nyquist sampling systems, specifically enhancing the modulated wideband converter by intentionally inducing aliasing to reduce sampling requirements without losing information.

Contribution

The paper proposes a novel aliasing method in the AMWC that decreases sampling rate and hardware complexity while maintaining lossless signal reconstruction, independent of PR signal speed.

Findings

Reduces the number of sampling branches needed

Lowers ADC sampling rate without information loss

Improves reconstruction performance with fixed hardware

Abstract

A modulated wideband converter (MWC) has been introduced as a sub-Nyquist sampler that exploits a set of fast alternating pseudo random (PR) signals. Through parallel sampling branches, an MWC compresses a multiband spectrum by mixing it with PR signals in the time domain, and acquires its sub-Nyquist samples. Previously, the ratio of compression was fully dependent on the specifications of PR signals. That is, to further reduce the sampling rate without information loss, faster and longer-period PR signals were needed. However, the implementation of such PR signal generators results in high power consumption and large fabrication area. In this paper, we propose a novel aliased modulated wideband converter (AMWC), which can further reduce the sampling rate of MWC with fixed PR signals. The main idea is to induce intentional signal aliasing at the analog-to-digital converter (ADC). In addition to the first spectral compression by the signal mixer, the intentional aliasing compresses the mixed spectrum once again. We demonstrate that AMWC reduces the number of sampling branches and the rate of ADC for lossless sub-Nyquist sampling without needing to upgrade the speed or period of PR signals. Conversely, for a given fixed number of sampling branches and sampling rate, AMWC improves the performance of signal reconstruction.