Full-Duplex MIMO Relaying: Achievable Rates under Limited Dynamic Range

Brian P. Day, Adam R. Margetts, Daniel W. Bliss, Philip Schniter

Abstract

This paper investigates the practical performance limits of full-duplex multiple-input multiple-output (MIMO) relaying between multi-antenna source and destination nodes. The primary challenge in such systems is self-interference (SI), where the relay's transmit signal overwhelms its receive circuitry. By explicitly modeling transmitter and receiver dynamic-range limitations (LDR) and channel estimation errors, the authors derive tight achievable rate bounds and propose a transmission scheme based on non-convex optimization. The study reveals a "critical INR" threshold beyond which the system must transition toward half-duplex behavior to maintain performance, providing a realistic framework for future full-duplex system design.

Introduction: The Challenge of Self-Interference

In an ideal scenario, a Full-Duplex (FD) relay would double spectral efficiency by transmitting and receiving simultaneously on the same frequency. However, real-world implementation is hindered by massive self-interference—often 100 dB stronger than the desired incoming signal. While digital cancellation techniques can mitigate some of this interference, they are fundamentally limited by the hardware's Limited Dynamic Range (LDR). Components such as Analog-to-Digital Converters (ADCs), Digital-to-Analog Converters (DACs), and power amplifiers introduce distortion that scales with signal power, creating a "noise floor" that cannot be removed through standard cancellation methods.

Modeling Hardware Imperfections

To address these limitations, the authors employ a signal model that treats hardware distortion as additive Gaussian noise with variance proportional to the signal power. This approach moves beyond the assumption of perfect hardware to account for the "DR Wall" created by limited resolution.

Detailed signal flow model for a full-duplex MIMO relay, incorporating transmitter/receiver distortion and channel estimation errors.

As shown in the model above, the relay's input and output are affected by distortion terms ($e_r$ and $c_r$). These terms represent the inherent "muck" of the hardware, making perfect cancellation impossible even if the transmitted signal is known exactly. The paper also accounts for Least-Squares (LS) channel estimation errors, showing how hardware limits degrade the accuracy of the "map" the relay uses to manage interference.

Optimization and Performance Bounds

The paper focuses on maximizing the end-to-end achievable rate for a Decode-and-Forward (DF) relay. This requires balancing the source-to-relay and relay-to-destination links, which is formulated as a "maximin" optimization problem. Because the mutual information terms are non-convex functions of the transmit covariance matrices, the authors developed a novel optimization approach combining a bisection search (to balance the links) with Gradient Projection (to find optimal transmit beam patterns).

They also propose the "Transmit Covariance Optimization with 2 periods and Interference Cancellation" (TCO-2-IC) scheme. This allows the relay to intelligently use its spatial degrees of freedom to "null" its own transmission, effectively trading off antennas used for data transmission for those used for interference suppression.

The Full-Duplex to Half-Duplex Transition

A key insight of the research is the existence of a "critical INR" (Interference-to-Noise Ratio). This threshold defines the point where the hardware distortion from self-interference becomes too severe to justify pure full-duplex operation.

Operating regions for full-duplex and half-duplex signaling as a function of SNR and INR, showing the critical INR threshold.

As illustrated in the figure above, when self-interference is low to moderate, full-duplex signaling provides massive gains. However, as the INR increases, the optimized system naturally "backs off," eventually behaving like a Half-Duplex (HD) system to avoid drowning in its own noise. The derived lower bounds for achievable rates are shown to be very close to theoretical upper bounds, especially as training length increases, validating the robustness of the proposed optimization framework.

Conclusion

This work provides a foundational look at the practical hardware constraints that prevent full-duplex systems from reaching their theoretical potential. By quantifying the spatial cost of managing self-interference and identifying the critical INR boundary, the authors offer a realistic roadmap for implementing MIMO relays in modern radio environments. The results emphasize that while full-duplex offers significant spectral gains, its benefits are strictly governed by the dynamic range of the underlying hardware.

Summary updated

Abstract

In this paper we consider the problem of full-duplex multiple-input multiple-output (MIMO) relaying between multi-antenna source and destination nodes. The principal difficulty in implementing such a system is that, due to the limited attenuation between the relay's transmit and receive antenna arrays, the relay's outgoing signal may overwhelm its limited-dynamic-range input circuitry, making it difficult---if not impossible---to recover the desired incoming signal. While explicitly modeling transmitter/receiver dynamic-range limitations and channel estimation error, we derive tight upper and lower bounds on the end-to-end achievable rate of decode-and-forward-based full-duplex MIMO relay systems, and propose a transmission scheme based on maximization of the lower bound. The maximization requires us to (numerically) solve a nonconvex optimization problem, for which we detail a novel approach based on bisection search and gradient projection. To gain insights into system design tradeoffs, we also derive an analytic approximation to the achievable rate and numerically demonstrate its accuracy. We then study the behavior of the achievable rate as a function of signal-to-noise ratio, interference-to-noise ratio, transmitter/receiver dynamic range, number of antennas, and training length, using optimized half-duplex signaling as a baseline.
BibTeX
@article{day2012full,
  author = {Brian P. Day and Adam R. Margetts and Daniel W. Bliss and Philip Schniter},
  title = {{Full-Duplex MIMO Relaying: Achievable Rates under Limited Dynamic Range}},
  journal = {IEEE J. Sel. Areas Commun.},
  year = {2012},
  url = {https://openreview.net/forum?id=wHs5c4cfZRuO},
}
Figure