Abstract
We study the potential employment of improper Gaussian signaling (IGS) in full-duplex cooperative settings with residual self-interference (RSI). IGS is recently shown to outperform traditional proper Gaussian signaling (PGS) in several interference-limited channel settings. In this work, IGS is employed in an attempt to alleviate the RSI adverse effect in full-duplex relaying (FDR). To this end, we derive a tight upper bound expression for the end-to-end outage probability in terms of the relay signal parameters represented in its power and circularity coefficient. We further show that the derived upper bound is either monotonic or unimodal in the relay's circularity coefficient. This result allows for easily locating the global optimal point using known numerical methods. Based on the analysis, IGS allows FDR systems to operate even with high RSI. It is shown that, while the communication totally fails with PGS as the RSI increases, the IGS outage probability approaches a fixed value that depends on the channel statistics and target rate. The obtained results show that IGS can leverage higher relay power budgets than PGS to improve the performance, meanwhile it relieves its RSI impact via tuning the signal impropriety.
Introduction
Full-duplex (FD) radio technology represents a significant leap for wireless networks, offering the potential to double spectral efficiency by allowing simultaneous transmission and reception on the same frequency. However, the practical implementation of FD is hindered by Residual Self-Interference (RSI). Despite advanced cancellation techniques, the relay's own transmitted signal often leaks back into its receiver, significantly degrading the signal-to-interference-plus-noise ratio (SINR). Traditionally, engineers use Proper Gaussian Signaling (PGS), which is optimal for noise-limited channels but performs poorly in interference-limited environments like FD relaying.
System Model
The researchers analyze a decode-and-forward (DF) relay system consisting of a source (S), a relay (R), and a destination (D). As illustrated in the system model below, the relay faces interference from its own transmissions (RSI) while attempting to receive the signal from the source.
Methodology: Improper Gaussian Signaling
The paper proposes replacing PGS with Improper Gaussian Signaling (IGS) at the relay. Unlike PGS, where the real and imaginary components of the signal are independent and have equal power, IGS breaks this circular symmetry. This "impropriety" is characterized by a circularity coefficient ($C_x$), ranging from 0 (Proper) to 1 (Maximally Improper), which allows for an additional degree of freedom in signal design.
The authors derived a tight upper bound for the end-to-end outage probability and demonstrated that this bound is either monotonic or unimodal with respect to $C_x$. This mathematical property ensures that the optimal circularity coefficient can be efficiently found using numerical methods like the bisection method, making it suitable for real-time applications.
Performance Analysis and Key Results
The core finding is the "interference immunity" provided by IGS. While PGS-based systems suffer from a performance floor or even total failure as RSI or relay power increases, IGS allows the system to maintain a stable outage probability.
As shown in the figure above, increasing the relay power budget in a PGS system eventually leads to a "breakeven point" where the increased RSI outweighs the benefits of a stronger signal to the destination (see the PGS-MPA curve). In contrast, IGS enables the relay to utilize higher power budgets effectively by tuning the signal's impropriety to suppress the impact of RSI. In extreme cases, the optimal strategy involves one-dimensional signaling ($C_x \to 1$), which essentially clears one dimension of the complex signal space for receiving the source's signal.
Conclusion
This work demonstrates that IGS is a powerful tool for overcoming the fundamental limitations of full-duplex relaying. By optimizing the signal shape, FDR systems can operate reliably even in high-interference environments. While implementing IGS requires more sophisticated transceiver designs and accurate channel state information, the gains in robustness and power efficiency are substantial. Future research may extend these concepts to MIMO systems and multi-relay networks to further exploit spatial and signal-space degrees of freedom.
Summary updated