Phased Array-Based Sub-Nyquist Sampling for Joint Wideband Spectrum Sensing and Direction-of-Arrival Estimation

TL;DR

This paper introduces a phased-array based sub-Nyquist sampling method for wideband spectrum sensing and DoA estimation, enabling accurate recovery without sparse spectrum constraints at reduced sampling rates.

Contribution

A novel sub-Nyquist sampling architecture using variable delays and CP decomposition for joint wideband spectrum sensing and DoA estimation, avoiding sparse spectrum assumptions.

Findings

Achieves near-CRB estimation accuracy with few samples

Does not require sparse spectrum constraints

Sampling rate exceeds the largest narrowband bandwidth

Abstract

In this paper, we study the problem of joint wideband spectrum sensing and direction-of-arrival (DoA) estimation in a sub-Nyquist sampling framework. Specifically, considering a scenario where a few uncorrelated narrowband signals spread over a wide (say, several GHz) frequency band, our objective is to estimate the carrier frequencies and the DoAs associated with the narrowband sources, as well as reconstruct the power spectra of these narrowband signals. To overcome the sampling rate bottleneck for wideband spectrum sensing, we propose a new phased-array based sub-Nyquist sampling architecture with variable time delays, where a uniform linear array (ULA) is employed and the received signal at each antenna is delayed by a variable amount of time and then sampled by a synchronized low-rate analog-digital converter (ADC). Based on the collected sub-Nyquist samples, we calculate a set of cross-correlation matrices with different time lags, and develop a CANDECOMP/PARAFAC (CP) decomposition-based method for joint DoA, carrier frequency and power spectrum recovery. Perfect recovery conditions for the associated parameters and the power spectrum are analyzed. Our analysis reveals that our proposed method does not require to place any sparse constraint on the wideband spectrum, only needs the sampling rate to be greater than the bandwidth of the narrowband source signal with the largest bandwidth among all sources. Simulation results show that our proposed method can achieve an estimation accuracy close to the associated Cram\'{e}r-Rao bounds (CRBs) using only a small number of data samples.