[Paper Review] Co-Designing Statistical MIMO Radar and In-band Full-Duplex Multi-User MIMO Communications -- Part I: Signal Processing

Jiawei Liu, Kumar Vijay Mishra+1 more|arXiv (Cornell University)|Jun 26, 2020

Radar Systems and Signal ProcessingEngineering62 references37 citations

TL;DR

The paper proposes a co-design framework for a statistical MIMO radar and IBFD MU-MIMO communications system sharing spectrum, jointly optimizing radar codes, UL/DL precoders, and receive filters via a compounded-and-weighted sum mutual information criterion.

ABSTRACT

We consider a spectral sharing problem in which a statistical (or widely distributed) multiple-input multiple-output (MIMO) radar and an in-band full-duplex (IBFD) multi-user MIMO (MU-MIMO) communications system concurrently operate within the same frequency band. Prior works on joint MIMO-radar-MIMO-communications (MRMC) systems largely focus on either colocated MIMO radars, half-duplex MIMO communications, single-user scenarios, omit practical constraints (clutter, uplink [UL]/downlink [DL] transmit powers, UL/DL quality-of-service, and peak-to-average-power ratio), or MRMC co-existence that employs separate transmit/receive units. The purpose of this and companion papers (Part II and III) is to co-design an MRMC framework that addresses all of these issues. In this paper, we propose signal processing for a distributed IBFD MRMC, where radar receiver is designed to additionally exploit the downlink communications signals reflected from a radar target. Extensive numerical experiments show that our methods improve radar target detection over conventional codes and yield a higher achievable data rate than standard precoders. The following companion paper (Part II) describes the theory and procedure of our algorithm to solve the non-convex design problem. The final companion paper (Part II) considers the case of multiple targets and examines the tracking performance of our MRMC system.

Motivation & Objective

• Motivate and address spectrum sharing between a widely distributed MIMO radar and an in-band full-duplex MU-MIMO communications system operating in the same band.
• Jointly design radar waveforms and communications precoders/receivers under practical constraints (UL/DL power, QoS, PAR).
• Develop an optimization framework that yields a monotonic convergence and improved performance for both radar detection and communications data rates.

Proposed method

• Define a compounded-and-weighted sum mutual information (CWSM) as the joint performance metric for radar and communications.
• Formulate a non-convex optimization problem that jointly optimizes radar code matrix A, UL/DL precoders {Pu,i[k],Pd,j[k]}, and linear receive filters, under UL/DL power, QoS, and peak-to-average-power-ratio constraints.
• Solve the non-convex CWSM problem using a combination of block coordinate descent (BCD) and alternating projection (AP) methods to obtain iterative solutions.
• Model the statistical MIMO radar (widely distributed antennas) and IBFD MU-MIMO communications with practical synchronization, channel, and interference considerations.
• Leverage discrete and continuous optimization steps to ensure fast, monotonic convergence within a few iterations.

Experimental results

Research questions

• RQ1 How can a statistical MIMO radar and IBFD MU-MIMO communications system be co-designed to share spectrum effectively?
• RQ2 What is an effective information-theoretic performance metric that jointly captures radar and communications objectives?
• RQ3 Can a practical optimization algorithm (BCD-AP) converge quickly while satisfying QoS and PAR constraints?
• RQ4 Do optimized radar codes and precoders improve radar detection and communications data rates compared to conventional designs?

Key findings

• Radar detection probability improves up to 13% over conventional radar codes at a given false alarm rate with the optimized design.
• Communications data rates improve up to 30% with the proposed precoder design compared to standard precoders.
• The proposed BCD-AP algorithm demonstrates monotonic convergence within a few iterations.
• The framework accounts for UL/DL transmit powers, UL/DL QoS (minimum rates), and peak-to-average-power ratio constraints in a unified design.
• Preliminary results (earlier conference version) focused on precoder design, with this work extending to joint radar and communications optimization.