Deep learning based methods have been increasingly applied to wireless localization and sensing. However, many existing methods suffer from poor generalization, which limits their adaptability to unseen environments. Additionally, retraining these models demands a substantial amount of labeled data, which is both time-consuming and labor-intensive. To address these challenges, we propose GenLoc, a generalization framework that enables accurate mitigation of ranging and angle estimation errors in unseen environments via feature disentanglement, without the need for additional data collection and online training. The key lies in learning fine-grained domain-invariant representations that are first extracted by minimizing feature distribution discrepancies across various domains and then refined through domain-invariant and domain-specific feature decoupling. Extensive experiments were conducted on real-world datasets, covering five distinct scenarios, each with five different obstacles, and a thorough comparison with several existing approaches illustrates that GenLoc improves average distance and angle estimation accuracy by more than 22% and 28%, respectively, while others exhibit negligible improvements or significant degradation.

This paper presents $\mathbf{OLinear}$, a $\mathbf{linear}$-based multivariate time series forecasting model that operates in an $\mathbf{o}$rthogonally transformed domain. Recent forecasting models typically adopt the temporal forecast (TF) paradigm, which directly encode and decode time series in the time domain. However, the entangled step-wise dependencies in series data can hinder the performance of TF. To address this, some forecasters conduct encoding and decoding in the transformed domain using fixed, dataset-independent bases (e.g., sine and cosine signals in the Fourier transform). In contrast, we propose $\mathbf{OrthoTrans}$, a data-adaptive transformation based on an orthogonal matrix that diagonalizes the series' temporal Pearson correlation matrix. This approach enables more effective encoding and decoding in the decorrelated feature domain and can serve as a plug-in module to enhance existing forecasters. To enhance the representation learning for multivariate time series, we introduce a customized linear layer, $\mathbf{NormLin}$, which employs a normalized weight matrix to capture multivariate dependencies. Empirically, the NormLin module shows a surprising performance advantage over multi-head self-attention, while requiring nearly half the FLOPs. Extensive experiments on 24 benchmarks and 140 forecasting tasks demonstrate that OLinear consistently achieves state-of-the-art performance with high efficiency. Notably, as a plug-in replacement for self-attention, the NormLin module consistently enhances Transformer-based forecasters. The code and datasets are available at https://github.com/jackyue1994/OLinear.

In this paper, we present a novel multi-hop link scheduling framework that utilizes the vision perception from cameras of the road-side unit (RSU) as well as cameras of the vehicle to support the large-capacity and reliable transmission of high-speed dynamic vehicle network. Specifically, we propose a vision based link state identification method to determine whether the communications links among RSU and different vehicles are blocked or connected. We firstly utilize the 3D detection technique to obtain the vehicle spatial distribution in surrounding environment. Then, the geometric calculation is adopted to accurately analyze the link states between RSU and different vehicles. Moreover, we design an environmental statistical information based low-complexity link scheduling method, and utilize the joint statistical distribution of the residual transmission distance and the residual multi-hop latency to optimize the total transmission latency. Simulation results show that the proposed vision based link state identification method significantly outperforms the exiting methods, and the proposed link scheduling method can approximately achieve the optimal performance as that from the exhaustive search method but with much less computation overhead.

The fifth-generation (5G) wireless communication signal is ideal for positioning by the user equipment (UE) due to its high precision, low cost, low latency, and easy integration. However, UE-side positioning faces significant challenges in complex electromagnetic environments. Obstacles that block the line-of-sight (LOS) path can severely impact localization accuracy. This paper proposes a deep learning architecture with two-stage inference to achieve high-precision positioning at the user side with multipath channels. For the first time, online learning is achieved without the need to retrain the neural network (NN), making it suitable for implementation on UE. Specifically, the variational inference theory is used to sense the environment and improve positioning accuracy using multipath information. This approach significantly reduces range error in non-line-of-sight (NLOS) channels. Furthermore, a new NN with a neural processes regressor (NPR) is proposed to learn the distribution of ranging bias directly from received signals. The proposed learning architecture and networks can be implemented without retraining, even under different circumstances and environments, making it more computationally efficient than most existing NNs and suitable for practical deployment. Simulation results demonstrate that the proposed approach outperforms conventional techniques with various channels.

Channel knowledge map (CKM) is viewed as a digital twin of wireless channels, providing location-specific channel knowledge for environment-aware communications. A fundamental problem in CKM-assisted communications is how to construct the CKM efficiently. Current research focuses on interpolating or predicting channel knowledge based on error-free channel knowledge from measured regions, ignoring the extraction of channel knowledge. This paper addresses this gap by unifying the extraction and representation of channel knowledge. We propose a novel CKM construction framework that leverages the received signals of the base station (BS) as online and low-cost data. Specifically, we partition the BS coverage area into spatial grids. The channel knowledge per grid is represented by a set of multi-path powers, delays, and angles, based on the principle of spatial consistency. In extracting these channel parameters, the challenges lie in strong inter-cell interference and non-linear relationships between received signals and channel parameters. To address these issues, we formulate the problem of CKM construction into a problem of Bayesian inference, employing a interference-activity prior model to characterize the path-loss differences of interferers. Under the Bayesian inference framework, we develop a hybrid message-passing algorithm for the interference-cancellation-based CKM construction. Based on the CKM, we obtain the joint frequency-space covariance of the user channel and design a CKM-assisted Bayesian channel estimator. The computational complexity of the channel estimator is substantially reduced by exploiting the CKM-derived covariance structure. Numerical results show that the proposed CKM provides accurate channel parameters at low signal-to-interference-plus-noise ratio (SINR) and that the CKM-assisted channel estimator significantly outperforms state-of-the-art counterparts.

In this paper, we consider the topic of multiple-input multiple-output (MIMO) detection for coded systems. Tree search based detectors are well-behaved but have high complexity, while low-cost random sampling based detectors may suffer from distribution mismatch and performance loss. To overcome these issues, we propose a data-driven algorithm, called trained parameter based path sampling (TPbPS), to optimize parameters of the sampling distribution for random sampling based detectors. A more robust approach is designed to avoid sampling repetition and missing by derandomizing the path sampling, where paths are determined by the trained sampling distribution as well as fixed and uniformly-distributed numbers. The error probability is derived for the proposed TPbPS, and it shows that the full diversity can be achieved by elaborately designing a set of sampling path for each quantized noise level. Further, a soft-TPbPS algorithm of computing reliable soft outputs for channel decoders is proposed, via applying end-to-end Bayesian optimization to learn parameters of sampling distribution that goes straight at the objective of optimizing the block error rate performance. Combining these techniques yields enhanced soft MIMO detection designs that non-trivially advance the state-of-the-art, and provide significant performance or complexity gains over traditional methods.

To support object detection and tracking as one of the main services of a mono-static integrated sensing and communications (ISAC), an ISAC node needs to operate in full-duplex mode, and therefore suffers from strong self-interference (SI) mainly due to mutual coupling (MC) between its antenna arrays. This paper studies beamforming (BF) designs at transmitter (Tx) and/or receiver (Rx) sides of ISAC node to support both multibeam transmissions for simultaneous communications and sensing, and also manage SI cancellation (SIC) over signaling bandwidth and suppress specific spatial reception angles. Provided that ISAC node has estimated SI channel only over a few frequency indices, a null-space projection (NSP) based BF designs at Tx-only, Rx-only, and joint Tx-Rx are studied to see which one is preferred from hardware-efficiency perspective. Moreover, a system level analysis with a proposed sensing frame structure is discussed to support reliable and periodic sensing of a region by choosing proper parameters while minimizing communications performance loss. Performance of SIC and effect of number of its constraints are discussed through numerical results. It is shown than NSP-based Tx-BF is sufficient for SIC and one can relax RxBF by only utilizing simple phase-shifters. Moreover, importance of having precise estimation of MC, and performance of object detection with and without SIC are discussed.

Precise positioning plays a critical role in beyond-fifth-generation ($\mathbf{B 5 G}$) wireless networks for a wide range of indoor and outdoor applications. Conventionally, positioning accuracy can be dramatically degraded in a high Doppler environment and it may also require complex signal processing technique to accurately estimate the positioning in harsh multipath channels. Thus, this paper studies a positioning reference signal with orthogonal time-frequency space (PRS-OTFS) to enhance multi-user positioning accuracy. Firstly, multi-user PRS are designed in delay-Doppler (DD) domain and then a simple DD-domain based algorithm is proposed to estimate time-of-arrivals (ToAs) which is mainly based on an efficient energy detection in DD domain. Performance superiority of the proposed DD-domain scheme compared to conventional time-frequency based estimators is also discussed. Finally, robustness of the ToA estimation algorithm is discussed through numerical results for various channel conditions and arbitrary non-integer normalized delay-and Doppler-spread.

Deep learning (DL) and artificial intelligence (AI) based channel estimation (CE) and multi-input and multi-output (MIMO) detection (MIMODet), have gained emerging interests in communication systems, aiming for leveraging the advantages of AI. However, it is a non-trivial task for AI based CE or MIMODet to outperform classical algorithms, especially when those are close to optimal. For instance, the conventional linear minimum mean-square-error (LMMSE) based CE, and QRdecomposition and M-algorithm (QRM) based MIMO detection, respectively. In this paper, we propose a novel AI receiver design that connects CE and MIMODet in a unified neural network (NN) architecture and adopts structural input features and offline-trainings. The CE utilizes a super-resolution based convolutional NN (SRCNN), and the MIMODet applies a QRMNet based on QRM detector enhanced graphical NN (GNN). Numerical results show that our AI receiver can achieve accurate CE and near-optimal block error rate (BLER) performance in tests of the fifth-generation new-radio ($5 \mathrm{G}-\mathrm{NR}$) system.

In the interference rejection combining (IRC) algorithm, the sample covariance matrices generally have to be stored before being utilized, which can be extremely costly due to the large volumes of matrices in applied slots or bandwidth, coupled with its difficult compression characteristics. In this paper, we design a compression algorithm which can significantly reduce the storage requirement while maintaining the performance of IRC. The core of our algorithm is to transform the matrix to be compressed into a diagonally dominant matrix via a suitable unitary transform, and to store the off-diagonal entries in floating point numbers with shorter bit-widths. Besides the storage, the computation expense of the proposed unitary transform is low. Theoretical analysis and numerical results are both provided to demonstrate the effectiveness of proposed algorithm.

In this paper, we consider the interference rejection combining (IRC) receiver, which improves the cell-edge user throughput via suppressing inter-cell interference and requires estimating the covariance matrix including the inter-cell interference with high accuracy. In order to solve the problem of sample covariance matrix estimation with limited samples, a regularization parameter optimization based on the minimum eigenvalue criterion is developed. It is different from traditional methods that aim at minimizing the mean squared error, but goes straight at the objective of optimizing the final performance of the IRC receiver. A lower bound of the minimum eigenvalue that is easier to calculate is also derived. Simulation results demonstrate that the proposed approach is effective and can approach the performance of the oracle estimator in terms of the mutual information metric.

In the real-time communication service over fifth-generation (5G) wireless network, e.g., FaceTime video call, the data packet loss and jitter, which increase the end - to-end packet delay and result in video stuck or glitching, influence the quality of experience (QoE) severely. In order to reduce packet delay and loss, an easily -configured enhancement scheme, called packet-level coding transmission optimization (PCTO), is proposed for user equipment, e.g., smart phone. In the proposed scheme, the data packets are firstly encoded by packet-level coding (PC) module in which some redundant packets are generated. Secondly, we differentiate the transmission priority for each packet by transmission optimization (TO) module, in which two principles are enforced. The first principle is that the transmission priority of data packet is higher than the redundant packet in order to overcome congestion cycle. The second principle is that the remaining uplink resource is allocated to code blocks uniformly. Simulation results show that the proposed scheme reduces 80 % packet loss and 22 % average delay at least and mean opinion score (MOS) can be improved by 33%. It can also be well adapted to different network conditions compared with existing schemes. Furthermore, the proposed scheme is compatible with the existing 5G standard. The procedures for configuring PCTO for an existing 5G network are also presented.

Achieving high-precision positioning holds paramount significance in the realms of intelligent manufacturing, the Internet of Things, and smart transportation. Nevertheless, attaining accurate labels and ensuring network generalization on mobile devices poses a formidable challenge. In this paper, we put forth an unsupervised and versatile wireless positioning methodology rooted in deep variational reinforcement learning (DVRL). The primary contributions of this work encompass the following: 1) the introduction of an approach to model wireless localization based on user-equipment (UE) as a partially observable Markov decision process (POMDP) and the treatment of this model within the DVRL framework; 2) the incorporation of a neural processes regressor (NPR) to derive ranging results from environmental cues, thus negating the need for model retraining; and 3) the alleviation of the challenge posed by the need for precise labels by deriving rewards from NPR's localization uncertainty, making use of unlabeled data. Our simulations affirm that this approach yields high-precision positioning without mandating precise labels or network retraining, surpassing conventional techniques in both positioning accuracy and generalization capabilities.

The maximum a posteriori (MAP) detector is widely regarded as the optimal one for soft multiple-input-multiple-output (MIMO) detection, but is infeasible to be used in a practical system due to its prohibitive cost. In this paper, we propose a soft MIMO detection theory approaching the MAP performance, which is called the MAP-moment (MAPM) algorithm. First, we fit the moments of the discrete joint posterior probability with a small number of observations. Then, with the help of deep unfolded Riemann-theta function, we can learn the true marginal posterior probability distribution from the aforementioned moments. Finally, bit log-likelihood ratios (LLRs) can be precisely derived by simple calculation. Simulation results indicate that the proposed algorithm can improve the performance significantly compared to State of the Art, close to the optimal MAP detector.

This paper proposes novel linear precoding algorithms for Multiple-Input Multiple-Output Bit-Interleaved Coded Modulation (MIMO-BICM) systems that maximize the achievable rate subject to power constraints. To overcome the nonlinear and nonconvex nature of the optimization problem, we rewrite the achievable rate in terms of the log-likelihood ratio (LLR) and introduce manifold-based gradient ascent (MGA) precoding and low-complexity non-iterative algorithms. Simulation results show significant gains in achievable rate and block error rate compared to existing techniques. Additionally, we extend our investigation to linear precoding with the constraint that the precoding matrix is selected from the codebook type-I adopted in Fifth-Generation New Radio (5G NR) networks. We propose heuristic algorithms that exploit the Kronecker and Discrete Fourier Transform (DFT) structure of the codebook and consider the singular vector decomposition (SVD) precoder as the optimal reference precoder. The traditional exhaustive search methods require a high complexity, especially for large codebook sizes. However, our proposed algorithms apply a combination of direct estimation and a low-dimensional search for deriving the indices, resulting in a reduced number of codebook precoder candidates. Simulation results show that our proposed low-complexity algorithms perform comparably to exhaustive search baselines.

Millimeter wave (mmWave) communication stands as a crucial technology for X-reality and cloud gaming due to its extensive bandwidth. Nevertheless, its user experience suffers from degradation caused by obstacles like humans and objects. Unlike base station (BS) solutions, user-side approaches offer greater flexibility in addressing blockage issues. This paper introduces an innovative beam management architecture named reference-signal-received-power (RSRP)-based blockage-prediction-aided compressive sensing (RSRP-BPCS). Unlike existing methods, our proposed algorithm accounts for real communication system constraints and protocol procedures, ensuring practicality. The key advantage lies in deriving channel information based on RSRP without requiring signal phase knowledge. Additionally, we present two novel strategies on the user side: a multiple-beam-sweeping procedure (MBSP) and a proactive blockage prediction approach. These techniques enable the monitoring of different beam pair links and early anticipation of blockage instances. To achieve rapid recovery, we design an explainable deep learning beam tracking scheme which is joint data and model-based to explore the compressive projection of the power spectrum. The best part is that our strategies are standard-compatible, demanding no supplementary side information, which significantly simplifies implementation with minimal complexity. Through extensive simulations, our proposed algorithm demonstrates exceptional performance in mitigating the impact of blockage, proving its efficacy and practicality.

In this paper, we propose a novel three-dimensional (3D) near-field beamforming (BF) design for Large Intelligent Surface (LIS). We firstly investigate the definitions of near-field and far-field of LIS, and derive the Fresnel near-field region where amplitudes variations are negligible but only phase variations worsen the harvested array-gains. We show that the Fresnel region which covers the majority part of near-field, can be enlarged by a factor of four when considering possible imperfectness from a conventional two-dimensional (2D) far-field BF. Therefore, it is of interest to design an analog 3D-BF that can recover array-gain losses in this region. Secondly, with a decomposition theorem we show that the optimal 3D-BF can be decomposed into a 2D far-field BF and a one-dimensional (1D) near-field BF. The 2D far-field BF compensates phase variations from mismatches in the azimuth and elevation angles, while the 1D near-field BF compensates remaining phases variations caused by distance differences from a user-equipment (UE) to different antenna-elements on LIS. Such a proposed “2D+1D” BF design reduces codebook-size significantly and is compatible with the existing far-field BF in the fifth-generation new-radio (5G-NR) system. Thirdly, we analyze an optimal codebook design for the 1D near-field BF, and show that with a small codebook it can perform close to optimal. Numerical results verify that the proposal is effective to recover array-gains in the near-field of LIS.

It is known that probabilistic constellation shaping (PCS) can provide a shaping-gain of 1.53dB asymptotically as signal-to-noise (SNR) increases. This is however, under ideal assumptions that the system operates in optimal sense and can achieve the Shannon capacity. In practice, the system can operate well below the capacity. In this paper, we propose a PCS-transceiver for the fifth-generation new-radio (5G-NR) evolution that supports quadrature-amplitude-modulation (QAM) constellations shaped with PCS, namely, PCS-QAM, in comparison to conventional uniform QAM constellations, namely, Uniform-QAM. We put a special interest in the throughput achieved under a stringent block-error rate (BLER) constraint, and validate the effectiveness of PCS under practical detection and decoding algorithms. We also analyze the properties of power-gain, entropy-loss, and peak-to-average power-ratio (PAPR), in connections to PCS. As a lower BLER does not necessarily indicate a higher throughput due to entropy-loss from PCS, we further prove that a necessary and sufficient condition for PCS-QAM to outperform Uniform-QAM in throughput is that the normalized entropy-loss with PCS-QAM is less than the BLER obtained with Uniform-QAM. Furthermore, we demonstrate that the PCS-transceiver is flexible in rate-adaptation without impacting the encoder, and can yield a better throughput-envelope in 5G-NR system.

Integrated sensing and communications (ISAC) has been deemed as a key technology for the sixth generation (6G) wireless communications systems. In this paper, we explore the inherent clustered nature of wireless users and design a multi-user based environment reconstruction scheme. Specifically, we first select users based on the estimation precision of channel’s multipath, including the line-of-sight (LOS) and the non-line-of-sight (NLOS) paths, to enhance the accuracy of environment reconstruction. Then, we develop a fusion strategy that merges communications signalling with camera image to increase the accuracy and robustness of environment reconstruction. The simulation results demonstrate that the proposed algorithm can achieve a remarkable sensing accuracy of centimeter level, which is about 17 times better than the scheme without user selection. Meanwhile, the fusion of communications data and vision data leads to a threefold accuracy improvement over the image only method, especially under challenging weather conditions like raining and snowing.

The high reliability and low latency requirements of multimedia services necessitate the design of more efficient link adaptation methods. In this paper, we introduce instantaneous channel state information (CSI) reporting, specifically designed for 5G video conferencing, and enhance the outer loop link adaptation based on soft Acknowledgement (Soft-ACK). We also formulate a resource allocation problem in 5G physical downlink shared channel (PDSCH) to balance the uplink and downlink traffic in compliance with the specified latency constraints. Our proposed scheme operates in a relatively straightforward manner. It outperforms conventional link adaptation methods re-garding Block-Level Error Rate (BLER) and effectively adheres to stringent latency constraints in video transmission simulations.

In this paper, we consider the topic of soft multipleinput multiple-output (MIMO) detection for coded systems. Conventional designs in this regard apply tree-based demappers that can achieve near-optimal bit error rate but can hardly guarantee the block error rate (BLER) performance without enough searching space. Therefore, we propose to improve the trained parameter based path sampling algorithm for computing reliable soft outputs. We first apply an autoencoder to train the sampling parameter under the concept of multi-label classification, aiming at minimizing the difference between the trained sampling distribution and well-predefined labels that represent the distribution of reliable maximum likelihood hypothesis and counter hypotheses. Further, end-to-end Bayesian optimization is utilized to learn the sampling parameter that goes straight at the objective of optimizing the BLER performance. The simulation results demonstrate that our algorithm yields an enhanced MIMO demodulation design that non-trivially advances the state-of-the-art, and can provide significant performance gains over traditional sphere decoders as well as being robust against channel variations.

The use of user equipment (UE) for positioning offers many benefits in indoor and urban areas, such as low latency and easy integration. However, acquiring accurate labels and achieving network generalization can be challenging. To address these issues, we propose a novel network architecture called variational neural processes (VNP), which is free from either precise labels or network retraining. VNP combines the advantages of variational inference and stochastic process regression theory. By generative model, environmental sensing is realized in weak-supervised manners. By neural processes, NLOS ranging bias can be always derived from environment-related information. Simulations demonstrate that our approach achieves high-precision positioning without requiring precise labels or network retraining, and it outperforms conventional techniques in terms of both positioning accuracy and generalization capability.

As 6G mobile services will become more immersive and user customized, users will pay more attention to their subjective experience. Energy efficiency and sustainability is also one of the foremost goals in 6G system design. However, traditional network or terminal optimization based on key performance indicator or key quality indicator is difficult to accurately reflect the subjective experience of users and may be superfluous in terms of energy consumption. Then, how to reduce the energy consumption without compromising user experience to ensure sustainability is a topic worthy of studying. In this paper, we propose a novel concept of end-to-end quality of experience (E2E-QoE) based collaborative optimization, and elaborate the challenges and designing aspects of implementing such a practical framework. In order to present a potential solution to ensure the above E2E-QoE based collaborative optimization, an E2E-QoE model is developed, which can provide uniform criterion for cross-layer optimization among over-the-top services, terminals, networks, etc. Further, we apply 5G real data into several use cases to verify the effectiveness and feasibility of the proposed framework.

Future wireless communication systems require increasing the data-rate, tremendously, given a limited bandwidth. One solution is to use super high order constellations in high signal to noise ratio (SNR) regimes. However, in practical systems, these constellations face many issues including hardware impairments and system non-linearities. In this paper we focus on the carrier frequency offset (CFO) and investigate its effects on the performance of an orthogonal frequency-division multiplexing (OFDM) system that applies a super quadrature amplitude modulation (super-QAM). Our analysis show that although super-QAM constellations are more sensitive to CFO than low-order QAM constellations, on the other hand, we can take advantage of the high SNR regime that super-QAM constellations are used in and obtain an accurate CFO estimate and consequently reduce error vector magnitude (EVM).

In this paper, reliable uplink transmission by 5G is investigated for task importance-oriented communication. A novel application-layer probabilistic constellation shaping (AL-PCS) scheme is proposed to improve the reliable transmission of data bits associated with more important task. Instead of physical layer (PL), the proposed scheme is implemented in the application layer (AL) of user equipment (UE), e.g., cell phone, in order to compatible with the existing 5G protocol. Firstly, the mapping table is constructed based on some related PL parameters. Then the output bitstream of applications are transformed to PCS bitstream by PCS module, in which a specific descrambling is preformed to counteract the scrambling operation in PL. Then after subsequent operations preformed by other layer, the obtained bitstream can be mapped to intended constellations with very high probability. Correspondingly, in the receiver, an PCS module is also implemented in application layer, in which the inverse operations are preformed to recover the original bitstream. For 256QAM and high code rate, the block error rate (BLER) performance can be improved by 13 dB.

The fifth-generation (5G) wireless communication is useful for positioning due to its large bandwidth and low cost. However, the presence of obstacles that block the line-of-sight (LOS) path between devices would affect localization accuracy severely. In this paper, we propose an online learning approach to mitigate ranging error directly in non-line-of-sight (NLOS) channels. The distribution of NLOS ranging error is learned from received raw signals, where a network with neural processes regressor (NPR) is utilized to learn the environment and range-related information precisely. The network can be implemented for online learning free from retraining, which is computationally efficient. Simulation results show that the proposed approach outperforms conventional techniques in terms of NLOS ranging error mitigation.

Soft demodulation of received symbols into bit log-likelihood ratios (LLRs) is at the very heart of multiple-input-multiple-output (MIMO) detection. However, the optimal maximum a posteriori (MAP) detector is complicated and infeasible to be used in a practical system. In this paper, we propose a soft MIMO detection algorithm based on marginal posterior probability statistics (MPPS). With the help of optimal transport theory and order statistics theory, we transform the posteriori probability distribution of each layer into a Gaussian distribution. Then the full sampling paths can be implicitly restored from the first- and second-order moment statistics of the transformed distribution. A lightweight network is designed to learn to recover the log-MAP LLRs from the moment statistics with low complexity. Simulation results show that the proposed algorithm can improve the performance significantly with reduced samples under fading and correlated channels.

In this paper, we propose a novel near-field beamforming (BF) design with a Large Intelligent Surface (LIS) that is implemented as a discretized 2D-array. We first investigate the definitions of the near-field and far-field regions, and determine the Fraunhofer distance of the LIS, which scales up linearly in the surface-area of the LIS. Hence, a user-equipment (UE) can enter the near-field of a LIS in practice. In addition to Fraunhofer distance, we further derive the Fresnel near-field region where both amplitude and angle variations are negligible, as long as the distance from the UE to the LIS is larger than a threshold, which only scales up linearly in the diameter of LIS. Therefore, in the majority region of near-field, only phase variations worsen the the quality of received signal and result in significant array-gain losses. Motivated by this observation, we further propose a two-step near-field BF design that can effectively recover the array-gain losses in Fresnel near-field, and is fully compatible with a conventional far-field BF.

Beam management under blockage has been challenging for millimeter wave communication relying on directional links. The baseline beam management protocols in 5G new radio (NR) are not able to predict and avoid the beam blockage timely, leading to significant link quality degradation and beam switching delays. In this paper, we propose a new dual- beam-sweeping-procedure (DBSP) scheme as well as a proactive blockage prediction and fast beam switching strategy at the user side. Specifically, we propose to only exploit existing resources in 5G NR standards and allocate them for DBSP, i.e., the regular beam sweeping procedure in 5G NR as well as an additional serving beam pair link (BPL) monitoring procedure. The DBSP then facilitates the user-side proactive blockage prediction. Once the beam blockage instance is predicted, users could determine candidate BPL earlier for fast beam switching so that the delays of beam switching can be reduced significantly. Compared to existing works, the proposed strategies are standard-compatible and require no additional side information, like location, map, or vision, making them easier to implement and of low complexity. Moreover, the proposed user-side strategies are more flexible than the base station (BS)-side methods, especially for beam switching/handover between BSs. Numerical results on both statistical and ray-tracing blockage channel models are provided to demonstrate the superiority of the proposed algorithms.

Beamforming design in non-line-of-sight (N-LOS) multipath channel is challenging for millimeter-wave (mmWave) user equipment (UE), especially when hardware imperfections introduce random phase distortions to the received signals. In this paper, we propose a novel analog beam codebook design algorithm for UE, called approximate correlation matrix (ACM). In the proposed algorithm, beamforming vector is designed according the reference signal received power (RSRP), without the need for exact phase information of received signals. In particular, we exploit the correlation among antennas from phaseless measurements, and derive the precoding/combining vector through a simple Fourier transform. Simulation results show that the proposed algorithms can achieve near-optimal performance with low complexity.

In this paper, we propose a novel approximate noise-whitening for multi-input multi-output (MIMO) detection via banded-inverse extension (BIE). If a noise covariance matrix origins from a Gauss-Markov process (GMP), its inverse is banded and the proposal is exact and yields no accuracy-losses. For general matrices to be inverted, the approximation-errors from inversion introduce a mismatched MIMO detection-model, which can cause performance-degradation. Hence, we develop an information-theoretic tool with generalized mutual information (GMI) to evaluate the impacts from approximation-errors on the final achievable rate. We show that the approximate noise-whitening method based on BIE not only minimizes the noise-distortion measured from the Kullback-Leibler (KL) divergence, but also asymptotically maximizes GMI of the MIMO system as signal-to-noise ratio (SNR) increases. Besides, the proposal also provides a unified framework for approximate matrix-inverse with an adjustable band-size that can tune the trade-off between complexity and accuracy.

In this paper, a low complexity MIMO detection algorithm based on path sampling is proposed to achieve near-optimal performance with limited number of paths. Conventional list sphere decoders (LSD) have good detection performance but at cost of high computation complexity, while the random sampling (RS) based MIMO detectors may suffer from inevitable performance loss and sample repetition. To overcome these issues, we first propose a data-driven method, called the trained parameter based path sampling (TPbPS), to optimize the parameter of sampling distribution, and design an approach to improve the generalization performance. Further, we take advantages of the well-trained sampling distribution and derandomize the RS to develop the robust parameter based path sampling (RPbPS), while all the sampling paths are determined by the trained sampling distribution as well as the fixed and uniformly-distributed numbers. Combining these techniques yields an enhanced MIMO detection design that non-trivially advances the state-of-the-art, and can provide significant performance and complexity gains over the traditional LSD and RS methods.

The rapid development of 5G communication technology has given birth to various real-time broadband communication services, such as augmented reality (AR), virtual reality (VR) and cloud games. Compared with traditional services, consumers tend to focus more on their subjective experience when utilizing these services. In the meantime, the problem of power consumption is particularly prominent in 5G and beyond. The traditional design of physical layer (PHY) receiver is based on maximizing spectrum efficiency or minimizing error, but this will no longer be the best after considering energy efficiency and these new-coming services. Therefore, this paper uses quality of experience (QoE) as the optimization criterion of the PHY algorithm. In order to establish the relationship between PHY and QoE, this paper models the end-to-end transmission from UE perspective and proposes a five-layer framework based on hierarchical analysis method, which includes system-level model, bitstream model, packet model, service quality model and experience quality model. Real data in 5G network is used to train the parameters of the involved models for each type of services, respectively. The results show that the PHY algorithms can be simplified in perspective of QoE.

Due to the fast deployment and commercial use of fifth generation (5G) communication systems, there is an increasing demand for higher uplink rates, and thus the deployment of more transmit antennas at user equipment (UE) becomes even more urgent. Nowadays, to better balance the uplink user experience and terminal cost, a feasible way to deploy larger number of transmit antennas at UE is to patch together multiple smaller radio frequency integrated circuits (RFICs), e.g., two RFICs of 2 transmit antennas (2T) will be used to implement 4T. However, this setup will induce a random phase difference between the two RFICs that is unknown at the transmitter. In this paper, we investigate the optimal uplink digital precoder design under such random phase imperfections. In terms of maximizing the average channel capacity, we find that the optimal precoder will be the eigen-vectors of an adjusted transmit correlation matrix. Block-diagonal precoders are also shown to be robust to random phase errors at the expense of some loss in degrees of freedom and channel capacity. Numerical simulations are provided to verify the effectiveness of the proposed uplink precoders under random phase impacts.

Compressive channel estimation is an effective way to reduce the number of measurements for fast beam tracking in millimeter wave (mmWave) communications. However in some practical scenarios, the phase of the received signal is difficult to measure, leading to challenges for fast beam tracking, especially in non-line-of-sight (NLoS) channels. In this paper, we propose a novel beam tracking algorithm under random phase offset scenarios using compressive sensing (CS). Unlike traditional algorithms based on complex signal measurements, the proposed algorithm could derive the channel information using reference signal received power (RSRP), without the need of knowledge of the phase. To recover the compressive channel with high precision and low complexity, we design a deep learning beam tracking scheme utilizing a complex-valued auto-encoder. Simulation results show that the proposed scheme can outperform traditional hierarchical search manner under blockage and rotation scenarios.

In this paper, we consider joint MIMO detection and LDPC decoding on a tripartite factor graph. Conventional method in this regard design a belief propagation (BP) detector that applies an matched filtering detection per layer by subtracting interferences from other layers, which is suboptimal. Further, the BP scheduling between detection and decoding is unoptimized, which makes the interaction less effective. To overcome these issues, we propose to enhance the BP based joint detection and decoding (BP-JDD) receiver with 4 different techniques: pre-filtering (PF), partial marginalization (PM), hybrid updating schedule (HUS) and damping. Among them, PF and PM can improve the BP detection under challenging conditions, such as correlated MIMO channels, high-order modulations, or high code-rates, while HUS and damping are effective to accelerate convergence and reduce the number of overall iterations for BP-JDD to succeed. Combining these 4 techniques yields an enhanced BP-JDD design that can provide significant performance gains over traditional turbo receivers, and is also robust against channel variations.

Orthogonal frequency division multiplexing (OFDM) is susceptible to carrier frequency offset (CFO), resulting in performance degradation in wireless communication. In this paper, we propose a log-likelihood ratio (LLR) optimization approach for residual CFO (called CFO-LLR), without the knowledge of exact CFO value. In such an approach, the variance of CFO is estimated through the measured signal-to-noise ratio (SNR), and the closed-form expression of LLR is derived. To deal with the mismatch between the supposed and real CFO distribution, a machine learning scheme is designed to learn the optimal LLR in practical scenarios with low complexity and flexible adaptability, called CFO-Net. Numerical results show that the proposed approach can achieve up to 4.6 dB gain over conventional demodulation scheme under typical CFO scenarios.

In order to reduce the downlink training overhead of mmWave communications, we propose a novel downlink beamforming strategy using the uplink sub-6GHz channel and downlink mmWave pilots that are sent from a few active antennas. Specifically, we design a novel dual-input neural network architecture, called FusionNet, to merge the sub-6GHz channel and the channel of a few active mmWave antennas. The proposed fusion model could intelligently adjust the attention paid (by the neural network) for sub-6GHz channel and mmWave channel by an attention mechanism. The output of the FusionNet represents the probability for each beam being the optimal one. We also propose an antenna selection model that can choose better active antennas to send the downlink pilots, in which the gradient of antenna selection vector is approximated by that of an antenna probability vector. Simulation results demonstrate the superior performance of the proposed strategy compared to the existing one that purely relies on the sub-6GHz information or compared to the shallow model that directly adds uniform pilots.