Machine Learning-Based Self-Interference Cancellation for Full-Duplex Radio: Approaches, Open Challenges, and Future Research Directions

Abstract

In contrast to the long-held belief that wireless systems can only work in half-duplex mode, full-duplex (FD) systems are able to concurrently transmit and receive information over the same frequency bands to theoretically enable a twofold increase in spectral efficiency. Despite their significant potential, FD systems suffer from an inherent self-interference (SI) due to a coupling of the transmit signal to its own FD receive chain. Selfinterference cancellation (SIC) techniques are the key enablers for realizing the FD operation, and they could be implemented in the propagation, analog, and/or digital domains. Particularly, digital domain cancellation is typically performed using modeldriven approaches, which have proven to be insufficient to seize the growing complexity of forthcoming communication systems. For the time being, machine learning (ML) datadriven approaches have been introduced for digital SIC to overcome the complexity hurdles of traditional methods. This paper reviews and summarizes the recent advances in applying ML to SIC in FD systems. Further, it analyzes the performance of various ML approaches using different performance metrics, such as the achieved SIC, training overhead, memory storage, and computational complexity. Finally, this paper discusses the challenges of applying ML-based techniques to SIC, highlights their potential solutions, and provides a guide for future research directions.

Bandit-Based Learning-Aided Full-Duplex/Half-Duplex Mode Selection in 6G Cooperative Relay Networks

“…More specifically, the closed-form expression of the pairwise-error probability was derived and optimal power allocation was conducted, based on the bisection method, under statistical source-relay ({S→R}) CSI at the relay, instantaneous relay-destination ({R→D}) CSI at the destination. Recently, ML-aided approaches for FD communications have been introduced, aiming to combat the impact of SI, mainly using ML data-driven algorithms digital SI cancellation to reduce the complexity of traditional methods in terms of CSI overheads [21]. Still, most of the existing ML-aided solutions rely on offline-trained ML algorithms for SI estimation over static SI channels.…”

6G: Technology Evolution in Future Wireless Networks

“…In [189], model-driven methods are executed for digital domain cancellation, which has shown to be inadequate to meet the increasing complexity of future communication systems. To get over the complexity barriers of conventional techniques, machine learning techniques have been created for digital Self-Interference Cancellation (SIC).…”