We consider a new paradigm for a user-centric network in millimeter-wave (mmWave) backhaul systems designed for high-speed vehicles. By deploying distributed MIMO along the travel routes of high-speed vehicles, inter-cell handovers and interference can be reduced while enhancing communication capacity by leveraging the diversity gain between distributed antennas (DA). However, challenges arise regarding the abrupt changes in Doppler frequency offset (DFO) and timing offset (TO) during DA switching. Frequent DA switching in high-speed environments degrades the demodulation performance. To address these issues, we propose a novel DFO and TO pre-compensation for seamless DA switching. The proposed method effectively estimates and pre-compensates DFO and TO by leveraging existing reference signal. Since the abrupt changes in DFO and TO are suppressed during switching, seamless switching can be achieved without requiring mobile terminal synchronization for each DA. Simulations demonstrate that the proposed method suppresses degradation of demodulation performance over a 160-millisecond interval, ensuring seamless DA switching.

This paper presents a novel radio resource allocation method for uplink scalable mmWave distributed MIMO (D-MIMO). The user-cluster-centric (UCC) partial minimum mean square error combining ($\mathrm{P}-\mathrm{MMSE}$) is a scalable reception scheme in large D-MIMO. However, the cluster-wise operation of the UCC approach causes performance degradation due to high intra- and inter-cluster interferences under high-user-density environments. In order to suppress the occurrence of both types of interferences, we propose an integrated radio resource and cluster allocation method that maximizes the performance of UCC-P-MMSE. The proposed method enables efficient suppression of the interference using two adjustable thresholds considering intra- and inter-cluster interferences. System-level simulations assuming an uplink mmWave D-MIMO system demonstrate that the proposed method improves a 5% -tile user throughput by $\mathrm{6 0 \%}$ compared to a conventional UCC-P-MMSE by avoiding the significant interference.

This paper proposes an outer loop link adaptation (OLLA) algorithm for massive multi-user multi-input multi-output (MIMO) systems that employs uplink multi-user detection (MUD) based on generalized approximate message passing (GAMP). The contribution aims to improve uplink system throughput performance for future beyond-fifth-generation mo-bile communication systems by designing a novel scheduler that can select spatially multiplexed user equipment (UE) devices and their modulation and coding schemes (MCSs), considering the high detection accuracy provided by the GAMP-based MUD. To achieve this, we propose an OLLA algorithm that accu-rately predicts the signal-to-interference and noise power ratio (SINR) that each UE can achieve after the GAMP-based MUD. Specifically, the proposed method can dynamically optimize the scheduler according to the iterative detection characteristics of GAMP by introducing a mechanism to correct the predicted SINR separately for each combination of spatially multiplexed UEs. System-level simulation results indicate that adjusting our OLLA algorithm achieves a 50% higher throughput than the conventional OLLA algorithm when applied to GAMP.

Linear minimum mean square error (LMMSE) filters are often utilized to achieve low-complexity multi-user detection (MUD) in uplink large multi-input multi-output (MIMO) systems. As the scale and density of MIMO systems grow towards truly massive setups, however, the LMMSE detection requiring high-dimensional matrix inversion operations becomes computationally expensive. As a promising approach to tackle this issue, the local LMMSE (LLMMSE) detector was proposed, where a contiguous block of beams can be selected for each user to construct the reduced beam-domain channels, assuming the use of digital beamforming at a base station (BS). A main issue is the performance degradation according to the angular spread of the received signal, due to the information loss induced by an excessive dimensionality reduction aiming at the computational reduction. To alleviate this issue, this paper proposes to selectively combine the information from the overlapped LLMMSE filters in the log-likelihood ratio (LLR) domain. In addition, this method is extended to probabilistic data association (PDA)-based iterative detection scheme for further enhancement of communication reliability. The efficacy of the proposed methods is demonstrated by simulation results in terms of the bit error rate (BER) performance and the computational cost.

We propose two full-digital receive beamforming (BF) methods for low-complexity and high-accuracy uplink signal detection via Gaussian belief propagation (GaBP) at base stations (BSs) adopting massive multi-input multi-output for open radio access network. In such scenarios, it is vital to reduce the cost of the BSs by limiting the bandwidth of fronthaul (FH) links, and the dimensionality reduction of the received signal based on receive BF at a radio unit is a well-known strategy to reduce the amount of data transported via the FH links. We clarify appropriate criteria for designing a BF weight considering the subsequent GaBP signal detection with the proposed methods: singular-value-decomposition-based BF and QR decomposition-based BF with the aid of discrete-Fourier-transformation-based spreading. Both methods enable dimensionality reduction without compromising the desired signal power by taking advantage of a null space of the channels. BF reduces correlations between the received signals in the BF domain, which improves the robustness of GaBP against spatial fading correlation. Simulation results indicate that the proposed methods improve detection capability while significantly reducing computation.

We consider large multi-user detection (MUD) via deep unfolding-aided belief propagation (BP) in coded multi-user MIMO (MU-MIMO) systems. A BP detector optimized (trained) by data-driven-tuning of embedded internal parameters achieves low-complexity and high-accuracy MUD while compensating practical imperfections. However, in actual implementation, these parameters should be optimized according to system parameters, e.g., modulation and coding scheme (MCS). In particular, when channel coding is used, it is vital not only to minimize the mean square error (MSE) but also to enhance the Gaussianity of the output log-likelihood ratio (LLR), in order to maximize the error correction capability of the subsequent soft-decision decoder. To that end, a novel loss function based on a weighted average of negentropy, which is a key measure to evaluate the Gaussianity, and MSE of the detector output is proposed. Simulation results show that the trainable Gaussian BP (T-GaBP) detector optimized with the proposed negentropy-aware loss function significantly improves the bit error rate (BER) performance of the decoder output and substantially outperforms the T-GaBP optimized with the typical MSE loss function.

This paper proposes a deep unfolding-aided belief propagation (BP) for large multi-user multi-input multi-output (MU-MIMO) detection under correlated fading channels. A BP-based detector is a well-known strategy for realizing large-scale MU detection (MUD) with low-complexity and high-accuracy. However, its convergence property is severely degraded under insufficient large-system conditions and spatial fading correlation among RX antenna elements. To compensate for this drawback, we design a trainable Gaussian BP (T-GaBP) having well-organized trainable internal parameters based on the BP structure. These parameters are optimized by the deep learning techniques in the signal-flow graph of unfolded GaBP; this approach is referred to as data-driven tuning. By training the parameters according to the system model, T-GaBP can maintain the high detection capability even in practical system configurations that differ from the ideal uncorrelated massive MIMO assumption. Numerical results show that the proposed detector improves the convergence property and achieves a comparable detection performance to the cutting-edge expectation propagation (EP) detector in correlated MUD, with a lower computational cost.