Limited Feedforward Waveform Design for OFDM Dual-Functional Radar-Communications

• Published in: IEEE transactions on signal processing Vol. 69; pp. 2955 - 2970
• Main Authors: Keskin, Musa Furkan, Koivunen, Visa, Wymeersch, Henk
• Format: Journal Article
• Language: English
• Published: New York IEEE 2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Backscattering; Communications systems; Covariance matrix; delay-Doppler ambiguity; Design optimization; dual-functional radar-communications; Estimation; Feedforward control; limited feedforward; Massive MIMO; Mathematical analysis; Matrix methods; OFDM; Parameter estimation; Radar; Receivers; Subcarriers; Time-frequency analysis; Tradeoffs; Transmitters; waveform design; Waveforms
• ISSN: 1053-587X; 1941-0476; 1941-0476
• DOI: 10.1109/TSP.2021.3076894

We consider the problem of time-frequency waveform design for an OFDM dual-functional radar-communications (DFRC) system that communicates with an OFDM receiver while simultaneously estimating target parameters using the backscattered signals. In particular, the goal is to achieve a favorable performance trade-off between radar and communications by optimizing subcarrier powers in a time-frequency region of interest. First, we focus on radar-optimal waveform design to minimize the Cramér-Rao bound (CRB) on delay-Doppler estimation subject to an integrated side-lobe level (ISL) constraint in the delay-Doppler ambiguity domain. Next, we investigate the problem of DFRC trade-off waveform design to optimize the communications rate under radar similarity constraint. Unlike the traditional DFRC systems which ignore feedforward overhead for conveying transmit waveform control information, we assume the existence of a low-rate feedforward channel between the DFRC transceiver and the communications receiver. Relying on the covariance matrix of linear minimum mean-squared-error (LMMSE) estimates of input symbols, we derive a novel communications metric as a function of both subcarrier powers and forwarded control information, and propose a joint waveform and control signaling optimization (JWCSO) strategy that leverages the sparsity and rank-one structure of DFRC waveforms within an alternating maximization framework. Simulation results show that the proposed JWCSO approach provides significant performance gains over the conventional feedforward-agnostic waveforms and achieves near-optimal radar-communications trade-off performance, reaching the boundary of the CRB-capacity region with only a limited feedforward information.