Spectrum sharing opportunities of full-duplex systems using improper Gaussian signaling

Mohamed Gaafar, Osama Amin, Walid Abediseid, Mohamed-Slim Alouini

Introduction

In the evolving landscape of wireless communications, spectrum scarcity remains a significant hurdle. Cognitive radio (CR) technology, specifically the "underlay" paradigm, offers a solution by allowing secondary users (SUs) to access licensed bands, provided they maintain the primary users' (PUs) quality of service (QoS). However, the emergence of Full-Duplex (FD) systems—where nodes transmit and receive simultaneously on the same frequency—introduces new complexities. FD nodes suffer from residual self-interference (RSI), which, when combined with interference from SUs, can severely limit spectrum sharing opportunities. Traditional approaches rely on Proper Gaussian Signaling (PGS), but this paper explores the untapped potential of Improper Gaussian Signaling (IGS) to enhance SU performance in FD environments.

System Model and Problem Formulation

The study considers a system comprising one FD PU pair and one SU pair. As illustrated in Figure 1, the FD PU nodes (Node 1 and Node 2) exchange information while simultaneously contending with their own RSI ($h_{11}, h_{22}$) and mutual interference ($h_{12}, h_{21}$). The SU transmitter aims to communicate with its receiver (link $g$) while its transmission creates interference at both PU nodes ($g_1, g_2$). Conversely, the PU transmissions interfere with the SU receiver ($h_1, h_2$).

System model of a secondary user (SU) pair sharing spectrum with a full-duplex primary user (PU) pair.

The primary challenge in this setup is that the PU receiver is already burdened by its own self-interference. Adding SU interference often pushes the PU's data rate below its required threshold, making traditional spectrum sharing unfeasible under standard signaling assumptions.

Methodology: Improper Gaussian Signaling

To overcome these limitations, the authors propose employing IGS for the SU. Unlike standard PGS, where the real and imaginary components of a signal are independent and have equal variance (circularly symmetric), IGS allows for correlation and unequal variances. This is characterized by the circularity coefficient ($C_x$), ranging from 0 (proper) to 1 (maximally improper).

The research formulates an optimization problem to maximize the SU's achievable rate ($R_s$) by jointly tuning its transmit power ($p_s$) and circularity coefficient ($C_x$). This optimization is subject to the PU's minimum rate requirement ($R_{0,p}$) and the SU's maximum power constraint. Due to the non-convex nature of the resulting mathematical problem, the authors developed Algorithm I, an efficient iterative method that identifies critical "breaking points" in the constraint space to find optimal solutions.

Key Results and Discussion

The proposed IGS scheme was evaluated through extensive Monte Carlo simulations using Rayleigh flat fading channel models. The results consistently demonstrate that IGS provides a significant performance boost over conventional PGS.

Achievable SU rate ($R_s$) vs. interference level ($\\bar{I}_s$) for various PU target rates ($R_{0,p}$).

As shown in Figure 2, IGS (blue lines) maintains a higher SU data rate compared to PGS (red dashed lines) across various interference levels ($\\bar{I}_s$) and PU target rates ($R_{0,p}$). The advantage of IGS is particularly pronounced when the PU's target rate is high or when the interference environment is challenging, often yielding a 1–4 dB improvement in achievable rates.

The core insight is that IGS enables "interference shaping." By adjusting the circularity coefficient, the SU can change the statistical structure of the interference it generates. This allows the SU to "steer" interference in a way that is less detrimental to the PU's specific receiver characteristics, effectively creating more "room" for its own transmission. Furthermore, IGS allows the SU to utilize its full power budget more effectively; while a PGS-based SU might have to lower its power to protect the PU, an IGS-based SU can maintain higher power by increasing its signal's "impropriety."

Conclusion and Future Directions

This work demonstrates that Improper Gaussian Signaling is a powerful tool for enabling spectrum sharing in complex Full-Duplex environments. By providing an additional degree of freedom—the circularity coefficient—IGS allows SUs to achieve higher data rates while strictly adhering to PU protection requirements.

However, practical implementation faces challenges. The current model assumes perfect Channel State Information (CSI), which is difficult to obtain in real-time. Additionally, IGS requires more complex transceiver hardware and signal processing than standard PGS. Future research should focus on robust IGS designs that account for CSI uncertainty and explore the scalability of this approach in multi-user cognitive networks.

Summary updated

Abstract

Sharing the licensed spectrum of full-duplex (FD) primary users (PU) brings strict limitations on the underlay cognitive radio operation. Particularly, the self interference may overwhelm the PU receiver and limit the opportunity of secondary users (SU) to access the spectrum. Improper Gaussian signaling (IGS) has demonstrated its superiority in improving the performance of interference channel systems. Throughout this paper, we assume a FD PU pair that uses proper Gaussian signaling (PGS), and a half-duplex SU pair that uses IGS. The objective is to maximize the SU instantaneous achievable rate while meeting the PU quality-of-service. To this end, we propose a simplified algorithm that optimizes the SU signal parameters, i.e, the transmit power and the circularity coefficient, which is a measure of the degree of impropriety of the SU signal, to achieve the design objective. Numerical results show the merits of adopting IGS compared with PGS for the SU especially with the existence of week PU direct channels and/or strong SU interference channels.
BibTeX
@inproceedings{gaafar2015spectrum,
  author = {Mohamed Gaafar and Osama Amin and Walid Abediseid and Mohamed-Slim Alouini},
  title = {{Spectrum sharing opportunities of full-duplex systems using improper Gaussian signaling}},
  booktitle = {PIMRC},
  year = {2015},
  url = {https://openreview.net/forum?id=HFBaIX1IO5},
}
Figure