Compressive Sampling Based UWB TOA Estimator

TL;DR

This paper introduces compressive sampling algorithms for UWB TOA estimation that operate at sub-Nyquist rates, achieving near-ML performance with reduced complexity and sampling requirements, especially when leveraging prior information.

Contribution

The paper presents novel compressive sampling-based TOA estimation algorithms, including a dictionary design and a method utilizing a-priori information, reducing sampling rates and computational complexity.

Findings

Achieves ML-like performance at 1/4 sampling rate at 25 dB SNR.

Reduces computational complexity compared to ML estimators.

Utilizes prior information to improve low SNR performance.

Abstract

This paper proposes two compressive sampling based time of arrival (TOA) estimation algorithms using a sub-Nyquist rate receiver. We also describe a novel compressive sampling dictionary design for the compact representation of the received UWB signal. One of the proposed algorithm exploits the a-priori information with regard to the channel and range of the target. The performance of the algorithms are compared against the maximum likelihood (ML) based receiver using IEEE 802.15.4a CM1 line of sight (LOS) UWB channel model. The proposed algorithm yields performance similar to the ML TOA estimation at high SNRs. However, the computational complexity and the sampling rate requirements are lesser compared to the ML estimator. Simulation results show that the proposed algorithms can match ML estimator performance with only 1/4-th the sampling rate at 25 dB SNR. We analyze the performance of the algorithm with respect to practical constraints like size of the holographic dictionary and sampling rates. We also propose a new algorithm which can exploit the a-priori information regarding the UWB channel and the geographical constraints on the target that may be available at the receiver. This algorithm can substantially boost the performance compared to the algorithm without a-priori information at low SNRs.