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. Self-interference 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 model-driven approaches, which have proven to be insufficient to seize the growing complexity of forthcoming communication systems. For the time being, machine learning (ML) data-driven approaches have been introduced for digital SIC to overcome the complexity hurdles of traditional methods. This article 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 article discusses the challenges of applying ML-based techniques to SIC, highlights their potential solutions, and provides a guide for future research directions.

In-band full-duplex (FD) technology has protruded as one of the most promising solutions for spectrum scarcity, by allowing users to transmit and receive simultaneously at the same center frequency. However, the FD systems suffer from a severe self-interference (SI) caused by the coupling between the transmit and receive antennas. Thereby, the potential of FD systems can not be attained without employing robust SI cancellation techniques. Traditionally, the SI is modeled using the polynomial-based cancelers, which are computationally expensive. Consequently, neural networks (NNs) have been recently introduced to model the SI with lower computational complexity. In this letter, a novel NN structure referred to as the dual neurons- $\ell $ hidden layers NN (DN- $\ell $ HLNN) is proposed. The DN- $\ell $ HLNN exploits two neurons in the first hidden layer to recognize the memory effect of the input and output samples with reduced complexity. The numerical simulations show that the DN- $\ell $ HLNN-based canceler significantly reduces the computational complexity and the memory requirements compared to the polynomial-and the existing NN-based cancelers while maintaining a similar non-linear cancellation performance.

Full-duplex (FD) systems have been introduced to provide high data rates for beyond fifth-generation wireless networks through simultaneous transmission of information over the same frequency resources. However, the operation of FD systems is practically limited by self-interference (SI), and efficient SI cancelers are sought to make the FD systems realizable. Typically, polynomial-based cancelers are employed to mitigate the SI; nevertheless, they suffer from high complexity. This article proposes two novel hybrid-layers neural network (NN) architectures to cancel the SI with low complexity. The first architecture is referred to as hybrid-convolutional recurrent NN (HCRNN), whereas the second is termed as hybrid-convolutional recurrent dense NN (HCRDNN).In the HCRNN, a convolutional layer is employed to extract the input data features using a reduced network scale. Moreover, a recurrent layer is then applied to assist in learning the temporal behavior of the input signal from the localized feature map of the convolutional layer. In the HCRDNN, an additional dense layer is exploited to add another degree of freedom for adapting the NN settings in order to achieve the best compromise between the cancellation performance and computational complexity. The complexity analysis of the proposed NN architectures is provided, and the optimum settings for their training are selected. The simulation results demonstrate that the proposed HCRNN and HCRDNN-based cancelers attain the same cancellation of the polynomial and the state-of-the-art NN-based cancelers with an astounding computational complexity reduction. Furthermore, the proposed cancelers show high design flexibility for hardware implementation, depending on the system demands.

In this paper, both linear and nonlinear features have been reviewed with linear support vector machine (SVM) classifier for neural seizure detection. The work introduced in the paper includes performance measurement through different metrics: accuracy, sensitivity, and specificity of multiple linear and nonlinear features with linear support vector machine (SVM). A comparison is performed between the performance of different combinations between 11 linear features and 9 nonlinear features to conclude the best set of features. It is found that some features enhance the detection performance greatly. Using a combination of 3 features of them, a linear SVM classifier detects seizures with sensitivity of 96.78%, specificity of 97.9%, and accuracy of 97.9%.

This paper proposes a hybrid quantum-classical machine learning (ML) approach for self-interference cancellation (SIC) in full-duplex transceivers. The proposed approach exploits a quantum long short-term memory (QLSTM) layer, integrated with a classical convolutional layer, to perform the non-linear SIC. QLSTM replaces the neural networks in the classical LSTM gates with variational quantum circuits, acting as feature extractors. Simulation results confirm the superiority of the proposed hybrid quantum-classical ML approach by achieving a significantly higher SIC performance than its fully classical counterpart at similar memory and computational requirements.

Due to the Internet of Things (IoT) requirements for a high-density network with low-cost and low-power physical (PHY) layer design, the low-power budget transceiver systems have drawn momentous attention lately owing to their superior performance enhancement in both energy efficiency and hardware complexity reduction. As the power budget of the classical transceivers is envisioned by using inefficient linear power amplifiers (PAs) at the transmitter (TX) side and by applying high-resolution analog to digital converters (ADCs) at the receiver (RX) side, the transceiver architectures with low-cost PHY layer design (i.e., nonlinear PA at the TX and one-bit ADC at the RX) are mandated to cope with the vast IoT applications. Therefore, in this paper, we propose the orthogonal shaping pulses minimum shift keying (OSP-MSK) as a multiple-input multiple-output (MIMO) modulation/demodulation scheme in order to design the low-cost transceiver architectures associated with the IoT devices. The OSP-MSK fulfills a low-power budget by using constant envelope modulation (CEM) techniques at the TX side, and by applying a low-resolution one-bit ADC at the RX side. Furthermore, the OSP-MSK provides a higher spectral efficiency compared to the recently introduced MIMO-CEM with the one-bit ADC. In this context, the orthogonality between the in-phase and quadrature-phase components of the OSP are exploited to increase the number of transmitted bits per symbol (bps) without the need for extra bandwidth. The performance of the proposed scheme is investigated analytically and via Monte Carlo simulations. For the mathematical analysis, we derive closed-form expressions for assessing the average bit error rate (ABER) performance of the OSP-MSK modulation in conjunction with Rayleigh and Nakagami-m fading channels. Moreover, a closed-form expression for evaluating the power spectral density (PSD) of the proposed scheme is obtained as well. The simulation results corroborate the potency of the conducted analysis by revealing a high consistency with the obtained analytical formulas.