OTFS-Based Joint Communication and Sensing for Future Industrial IoT

TL;DR

This paper introduces an OTFS-based joint communication and sensing scheme for industrial IoT, enhancing reliability, energy efficiency, and sensing accuracy while reducing costs and spectrum usage.

Contribution

It develops a novel echo pre-processing method for OTFS sensing and optimizes system parameters to improve target parameter estimation accuracy.

Findings

Proposed method approaches maximum likelihood estimator performance.

Achieves high sensing accuracy with low computational complexity.

Demonstrates robustness across various system parameters.

Abstract

Effective wireless communications are increasingly important in maintaining the successful closed-loop operation of mission-critical industrial Internet-of-Things (IIoT) applications. To meet the ever-increasing demands on better wireless communications for IIoT, we propose an orthogonal time-frequency space (OTFS) waveform-based joint communication and radio sensing (JCAS) scheme -- an energy-efficient solution for not only reliable communications but also high-accuracy sensing. OTFS has been demonstrated to have higher reliability and energy efficiency than the currently popular IIoT communication waveforms. JCAS has also been highly recommended for IIoT, since it saves cost, power and spectrum compared to having two separate radio frequency systems. Performing JCAS based on OTFS, however, can be hindered by a lack of effective OTFS sensing. This paper is dedicated to filling this technology gap. We first design a series of echo pre-processing methods that successfully remove the impact of communication data symbols in the time-frequency domain, where major challenges, like inter-carrier and inter-symbol interference and noise amplification, are addressed. Then, we provide a comprehensive analysis of the signal-to-interference-plus-noise ratio (SINR) for sensing and optimize a key parameter of the proposed method to maximize the SINR. Extensive simulations show that the proposed sensing method approaches the maximum likelihood estimator with respect to the estimation accuracy of target parameters and manifests applicability to wide ranges of key system parameters. Notably, the complexity of the proposed method is only dominated by a two-dimensional Fourier transform.