Intelligent Non-Orthogonal Beamforming With Large Self-Interference Cancellation Capability for Full-Duplex Multiuser Massive MIMO Systems

Requiring a single frequency band for transmission, full-duplex (FD) systems can ideally double the spectral efficiency. However, in addition to the information signal, radiated from user's own transmit antenna generates a strong self-interference (SI). To eliminate effects of SI, active and passive cancellation techniques are used in the literature. However, omitted in these studies is the fact that the time dispersion in the SI channel may cause an error floor. In this paper, it is verified that the SI channel is time dispersive via measurements. Later, we present an analysis in order to determine the error performance. Following these results, a fractionally spaced SI canceler is proposed to eliminate the error floor. We verify our analytical results via simulations, showing that the proposed SI canceler can critically enhance the performance of FD systems.

Underwater acoustic (UWA) communication suffers from the limited available bandwidth of acoustic links. We consider in-band full-duplex (FD) communication in UWA channels, which has the potential of doubling the channel capacity. The key challenge of FD operation is the strong self-interference (SI) introduced by the near-end transmission. For a time-invariant UWA channel, the SI can be efficiently cancelled using an adaptive filter. In practice, the time-varying reflections particularly introduced by the moving sea surface limit the cancellation performance. In this paper, we propose a two-stage SI cancellation scheme with a transceiver structure comprising one projector and two hydrophones. In the first stage, the stable SI path from each hydrophone is cancelled using a digital canceller based on the recursive least-squares (RLS) algorithm with dichotomous coordinate descent (DCD) iterations. In the second stage, the scheme tries to cancel the residual SI using an RLS-DCD-based adaptive beamformer. The beamformer combines the residual signals at the hydrophones in such a way to minimize the final residual SI. The cancellation performance of the proposed scheme is investigated using the Waymark simulator and with lake experiments. As demonstrated by the simulation results, the SI can be efficiently cancelled with the proposed two-stage SI cancellation scheme. The experimental results also demonstrate the capability of the adaptive beamformer in reducing the residual SI after digital cancellation.

In this paper, we investigate low-complexity user scheduling schemes for the downlink (DL) massive multiuser multiple-input multiple-output (MU-MIMO) system, where an M-antenna (M is very large) base station (BS) serves K (K > 2) N-antenna (N > 1) users with linear transceivers. Establishing the theoretical foundation of our scheduling schemes, we first investigate the asymptotic sum-rates of this system with both singular value decomposition (SVD) and regularized zero-forcing (RZF) precoders. In particular, we assume that both M and N go to infinity with a bounded ratio to perform the theoretical derivations with random matrix theory (RMT). In addition, the influence of imperfect channel estimation is also considered. Interestingly, the analysis indicates the existence of an optimal number of served users K <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sup> in terms of asymptotic sum-rate, which motivates our low complexity user scheduling schemes. Unlike the conventional schemes with complicated computations and instantaneous channel state information, our user scheduling schemes are based on semi-analytical solutions with only statistic CSI. Simulation results validate our theoretical derivations and show the advantages of our schemes in the massive MU-MIMO system.

The main aim of massive multiuser multiple-input multiple-output (MU-MIMO) system is to improve the throughput and spectral efficiency in 5G wireless networks. The performance of MU-MIMO system is severely influenced by inter-antenna interference (IAI) and multiuser interference (MUI). The IAI occurs due to space limitations at each user terminal (UT) and the MUI is added when one UT is in the vicinity of another UT in the same cellular network. IAI can be mitigated through a precoding scheme such as singular value decomposition (SVD), and MUI is suppressed by an efficient multiuser detection (MUD) schemes. The maximum likelihood (ML) detector has optimal performance; however, it has a highly complex structure and involves the need of a large number of computations especially in massive structures. Thus, the neighborhood search-based algorithm such as likelihood ascent search (LAS) has been found to be a better alternative for mitigation of MUI as it results in near optimal performance with low complexity. Most of the recent papers are aimed at eliminating either MUI or IAI, whereas the proposed work presents joint SVD precoding and LAS MUD to mitigate both IAI and MUI. The proposed scheme can achieve a near-optimal performance with smaller number of matrix computations.

This paper presents a kernel-based adaptive filter that is applied for the digital domain self-interference cancellation (SIC) in a transceiver operating in full-duplex (FD) mode. In FD, the benefit of simultaneous transmission and receiving of signals comes at the price of strong self-interference (SI). In this work, we are primarily interested in suppressing the SI using an adaptive filter namely adaptive projected subgradient method (APSM) in a reproducing kernel Hilbert space (RKHS) of functions. Using the projection concept as a powerful tool, APSM is used to model and consequently remove the SI. A low-complexity and fast-tracking algorithm is provided taking advantage of parallel projections as well as the kernel trick in RKHS. The performance of the proposed method is evaluated on real measurement data. The method illustrates the good performance of the proposed adaptive filter, compared to the known popular benchmarks. They demonstrate that the kernel-based algorithm achieves a favorable level of digital SIC while enabling parallel computation-based implementation within a rich and nonlinear function space, thanks to the employed adaptive filtering method.

Massive multiple-input multiple-output (MIMO) technology for wireless communication is considered a key technology for the future fifth-generation (5G) communication networks. To reduce the propagation loss of channel, MIMO systems with analog or digital hybrid precoders and combiners (HPC) architecture have been used. This review paper presents the various types of analog and digital HPC designs for the massive MIMO communication systems that were investigated within the last decade. The reviewed systems are useful for both single-user (SU) and multi-user MIMO (MU-MIMO) systems. The key parameter of the study was spectral efficiency (SE), signal-to-noise ratio (SNR), signal-to-interference pulse noise ratio (SINR), system sum-rate (SR) performance, computational complexity (CC), co-channel interference (CCI), inter-primary base station interference (IPBSI), multiuser interference (MUI), effective desired channel gain (EDCG), and minimum mean square error (MMSE).

Acquiring accurate channel state information (CSI) is essential for enabling reliable and efficient wireless transmission and reception. However, CSI is inherently stochastic, high-dimensional, and time-varying, which makes its acquisition particularly challenging. Motivated by the success of deep learning (DL) across many data modalities, this thesis explores DL-based solutions for CSI estimation, compression, and prediction.First, we study CSI estimation in full-duplex (FD) multiple-input multiple-output (MIMO) systems, where strong self-interference (SI) complicates channel acquisition. To reduce the pilot and computational burden of estimating both SI and user channels, we propose a pilot-sharing strategy together with a convolutional neural network that jointly estimates these channels.We further introduce a neural mapping that enables CSI acquisition at the transmit chain.Second, we investigate DL–based CSI compression and its limited robustness under distribution shifts. To address this issue, we adopt a full-model fine-tuning while explicitly accounting for model update signaling overhead. Specifically, we employ a spike-and-slab prior to promote sparsity in the model updates and fine-tune the pretrained network using a rate–distortion objective regularized by the update bit rate.Third, we tackle CSI prediction using a diffusion-based generative framework. The method consists of a temporal encoder that extracts latent features from past CSI and a diffusion generator that synthesizes future CSI. We also study a simplified encoder-free design to reduce latency, compare autoregressive and sequence-to-sequence inference, and explore multiple architectures for both temporal encoding and diffusion generation.

Full-duplex (FD) communication can potentially double spectral efficiency by allowing simultaneous transmission and reception over the same frequency. However, strong self-interference (SI) between the co-located transmitter (Tx) and receiver (Rx) frequently undermines the system performance. This paper introduces a two-stage beamformer optimization strategy to improve Tx-Rx beam-level isolation. First, a neural network model, trained on a dataset generated by particle swarm optimization, predicts optimal nulling angles for both downlink and uplink beamformers based on user locations. Next, the beam steering angle is perturbed within a predefined search window to further minimize SI. Validation using measured SI channels from an FD mMIMO testbed demonstrates an average Tx-Rx beam-level isolation of 95.7 dB, representing a 29.5 dB improvement, with less than a 3 dB trade-off in desired signal gain. Across all tested user pairs, beam-level mutual coupling remains better than −66.8 dB, and 95.9% of tested beam pairs achieve over 15 dB of additional isolation. The relationship between the allowable degradation in desired gain and the achievable Tx-Rx isolation enhancement is thoroughly examined.

In recent years, the utilization of In-Band Full-Duplex (IBFD) technology has gained considerable attention as an effective solution to address limited spectrum resources. However, achieving an IBFD system necessitates the utilization of both passive suppression and active cancellation techniques to completely eliminate strong self-interference (SI) signals in the local receiver. This research paper aims to tackle this challenge by introducing the Multi-Domain Intelligent Cancellation (MDIC) method, which integrates analog-domain and digital-domain cancellation techniques through comprehensive nonlinear system modeling. Moreover, this study presents an extensive analysis of the underlying principles of MDIC, while evaluating the limits of SI suppression performance under the assumption of perfect channel estimation and no phase noise. The channel parameter estimation is conducted using the time-domain least squares method, without considering the effects of Gaussian white noise and phase noise. To enhance the performance of SI cancellation (SIC), a deep learning network (DLN) is introduced in the digital domain. Simulation results demonstrate that the proposed MDIC method achieves a substantial SIC performance of approximately 60 dB, which closely approaches the theoretical limit. Furthermore, experimental data indicates that the proposed deep learning approach exhibits superior SIC performance and broader adaptability to broadband signals compared to traditional methods.

In this paper, a novel full-duplex non-coherent (FD-NC) transmission scheme is developed for massive multiple-input multiple-output (mMIMO) systems using analog beamforming (ABF). We propose to use a structured Grassmannian constellation for the non-coherent communications that does not require channel estimation. Then, we design the transmit and receive ABF via the slow time-varying angle-of-departure (AoD) and angle-of-arrival (AoA) information, respectively. The ABF design targets maximizing the intended signal power while suppressing the strong self-interference (SI) occurred in the FD transmission. Also, the proposed ABF technique only needs a single transmit and receive RF chain to support large antenna arrays, thus, it reduces hardware cost/complexity in the mMIMO systems. It is shown that the proposed FD-NC offers a great improvement in bit error rate (BER) in comparison to both half-duplex non-coherent (HD-NC) and HD coherent schemes. We also observe that the proposed FD-NC both reduces the error floor resulted from the residual SI in FD transmission, and provides lower BER compared to the FD coherent transmission.

In this paper, linear precoding (LP) algorithm has been studied to improve the bit error rate (BER) performance of large scale multiple input multiple output (MIMO) system. A new technique of enhancing the BER performance of linear precoding algorithm without complexity in hardware and mathematical calculations is proposed and it is used on a traditional LP algorithms like a Maximum Ratio Transmission (MRT), a Zero-Forcing (ZF), a Regularized Zero-Forcing (RZF), and a Minimum Mean Square Error (MMSE) for Millimeter wave MU-MIMO communication system. A new technique is also used to mitigate the BER gab problem between (ZF, MMSE) and (RZF, MMSE) linear precoding algorithms. The result shows that with using the new technique, the BER performance of modified LP significantly outperform their traditional LP by 10−5 when using 4 element of BS antenna, 10−6 when using 16 element of BS antenna and mitigate the BER gab problem at different signal to noise ratio (SNR).

We assume a full-duplex (FD) cooperative network subject to hostile attacks and undergoing composite fading channels. We focus on two scenarios: \textit{a)} the transmitter has full CSI, for which we derive closed-form expressions for the \textit{average secrecy rate}; and \textit{b)} the transmitter only knows the CSI of the legitimate nodes, for which we obtain closed-form expressions for the \textit{secrecy outage probability}. We show that secure FD relaying is feasible, even under strong self-interference and in the presence of sophisticated multiple antenna eavesdropper.

In this paper, we investigate the energy harvesting (EH) technique and accordingly design transceivers for a K link multiple-input multiple-output interference channel. Each link consists of two full-duplex (FD) Internet of Things (IoT) nodes exchanging information simultaneously in a bi-directional communication channel. All the nodes suffer from interference, in particular strong self-interference and inter-node interference, due to operating in FD mode and simultaneous transmission at each link, respectively. Further, we divide the received signal at each node into two parts. While one part of the signal is used for information decoding, the other part is used for EH. We jointly design the transmit and receive beamforming vectors and receiver power splitting ratios by minimizing the total transmission power of the system, subject to both signal-to-interference-plus-noise ratio and EH threshold constraints. Furthermore, the case of multiple-input single-output interference channel is also included for the sake of comparison. We also revisit the above problems for the case when the available channel state information (CSI) at the transmitters is imperfect, where the errors of the CSI are assumed to be norm bounded. Simulation results show that the EH technique can harvest enough energy to support power consumption limited IoT devices by aiding in recharging their respective batteries.

The performance of the hybrid beamforming system based on the full-connected structure is close to that of the full-digital one in millimeter-wave (mmWave) massive multiple-input multiple-output (MIMO) systems. However, due to the high complexity of circuit wiring in practical applications, it has been greatly restricted. This paper investigates the hybrid (analog/digital) beamforming design of a multiuser mmWave massive MIMO system with sub-connected structure. For this system, we focus on a sum-rate maximization probelm. Considering the joint design of the receiver, and transmitter, a two-stage design method is adopted. In the analog stage, the analog beamformer, and combiner are designed by the proposed piecewise dual joint iterative approximation (PDJIA) method, which not only can obtain closed-form solutions but also has linear property. Based on the criterion of avoiding the loss of information, the problem of digital beamforming is solved by exploiting the baseband piecewise successive approximation (BPSA) method, which can effectively increase the number of users served. Simulation results show that the proposed scheme outperforms the state-of-the art MIMO hybrid beamforming design schemes, and more accurately approximates to that of the full-digital system.

Massive multiple-input multiple-output (MIMO) achieves enhanced data rates with linear precoding schemes, making it an interesting PHY scheme for future mobile communications systems. The deployment of a high number of antennas facilitates reliable communication to many users in a multi-user MIMO system. However, reliable low-latency communication is only maintained, if all mobile users are served simultaneously in a multi-user MIMO fashion. In case of high inter-user correlation, for example, when two users are spatially close to each other, the user rate with linear precoding decreases. In the worst case, either of the two users is dropped by the scheduler in order to maintain a certain rate to at least one of them. We aim to improve a massive MIMO system’s reliability while keeping the number of antennas constant. To decrease user correlation of closely spaced users, we design an antenna array with a large aperture. We consider designing the antenna array by array thinning, employing a genetic algorithm. This leads to a nonuniform spacing of antennas. We show that the proposed array design maintains a certain user rate more reliably for linear precoding schemes compared to a uniform linear array.