In energy harvesting (EH) Internet of Bodies (IoB) networks, maintaining information freshness is critical for time-sensitive applications such as electrocardiogram (ECG) monitoring. Timely transmissions are essential for reliable ECG analysis, which requires complete and synchronous data from all sensors, as each ECG lead is derived from potential differences between paired sensors. To this end, we leverage model-free deep reinforcement learning (DRL) to learn adaptive scheduling policies that minimize the age of information (AoI) in dynamic and intermittent energy arrival conditions. Specifically, in a human body communication (HBC)-enabled EH-ECG network, wearable ECG sensors transmit data to a central wearable hub acting as the scheduler. The hub implements the DRL scheduling algorithm that selects sensor transmissions based on their battery and AoI levels. Furthermore, we propose an ECG synthesis framework for a 3-lead ECG that mitigates the impact of EH scarcity by reconstructing missing leads from data transmitted by active sensors, thereby ensuring complete signal availability at the hub. Simulation results show that the proposed solution reduces AoI by up to 57% compared to the greedy myopic policy and up to 96% relative to a baseline learning policy.

Reconfigurable intelligent surfaces (RIS) have emerged as a transformative technology for next-generation wireless networks, offering unprecedented control over radio propagation environments. However, their passive nature and ease of deployment introduces security vulnerabilities that remain largely unexplored. This paper investigates a spoofing attack where a malicious RIS strategically manipulates its reflection coefficients to impersonate a legitimate RIS, thereby deceiving the base station (BS) and gaining unauthorized network access. To counter this threat, we propose a novel authentication framework that formulates the detection problem as a data-driven binary classification task, leveraging conditional generative adversarial networks (cGAN). The framework employs a U-Net-based generator to synthesize realistic attack scenarios during training, while the discriminator serves as a lightweight authenticator enabling robust authentication without requiring apriori knowledge of attacker strategies. Through extensive simulations across diverse attack scenarios, including co-located and correlated configurations, we demonstrate that the trained discriminator achieves 96.4% detection accuracy against malicious RIS attackers positioned near the BS (co-located) and maintains 86.2% accuracy under correlated attack conditions.

Although massive multi-input multi-output (mMIMO) systems offer higher directivity gains to mitigate propagation losses at millimeter wave (mmWave) frequencies, they present challenges in channel state information (CSI) acquisition and beam alignment, particularly in highly mobile vehicle-to-vehicle (V2V) scenarios with short channel coherence times. To address these issues, we propose a target-in-the-loop beam tracking approach that leverages GPS data and Long Short-Term Memory (LSTM) networks to select beams from predefined beamforming codebooks. By transforming GPS data into relative coordinates and extracting features such as relative velocity and orientation, our model predicts future beam states up to 500 ms in advance, enabling proactive beam selection and blockage avoidance. Using the DeepSense V2V dataset, our method achieves up to 7.2 dB power loss reduction and a 32% improvement in top-5 accuracy compared to a linear interpolation baseline. This approach highlights the potential of integrating GPS data and machine learning to enhance beam tracking in dynamic V2V mmWave networks.

Accurate angle of arrival (AoA) and angle of departure (AoD) estimation underpins spatial precision, efficient beamforming, and the overall Quality of Service (QoS) and Quality of Experience (QoE) of integrated sensing and communication (ISAC). This paper introduces a generative AI (GenAI) framework that synthesizes the complete set of multipath 2D AoA/AoD parameters directly from transmitter (TX) and receiver (RX) locations, overcoming the limitations of traditional model-based and current data-driven learning methods. The proposed classifiers-guided conditional generative adversarial network (CG-CGAN) offers a system-agnostic approach that removes the need for pilot signals and operates effectively under challenging coherent multipath conditions. Its three-stage architecture integrates classification and conditional generative modeling to jointly infer line-of-sight (LoS) status, number of propagation paths, and angular parameters. Simulations on the DeepMIMO dataset demonstrate over 99.5% classification accuracy and 92% angle generation accuracy, significantly outperforming classical techniques while reducing computational complexity and enhancing the QoS/QoE of ISAC-enabled systems.

With the deployment of large antenna arrays at high-frequency bands, future wireless communication systems are likely to operate in the radiative near-field. Unlike far-field beam steering, near-field beams can be focused on a spatial region with a finite depth, enabling spatial multiplexing in the range dimension. Moreover, in the line-of-sight MIMO near-field, multiple spatial degrees of freedom (DoF) are accessible, akin to a scattering-rich environment. In this paper, we derive the beamdepth for a generalized uniform rectangular array (URA) and investigate how the array geometry influences near-field beamdepth and its limits. We define the effective beamfocusing Rayleigh distance (EBRD), to present a near-field boundary with respect to beamfocusing and spatial multiplexing gains for the generalized URA. Our results demonstrate that under a fixed element count constraint, the array geometry has a strong impact on beamdepth, whereas this effect diminishes under a fixed aperture length constraint. Moreover, compared to uniform square arrays, elongated configurations such as uniform linear arrays (ULAs) yield narrower beamdepth and extend the effective near-field region defined by the EBRD. Building on these insights, we design a polar codebook for compressed-sensing-based channel estimation that leverages our findings. Simulation results show that the proposed polar codebook achieves a 2 dB NMSE improvement over state-of-the-art methods. Additionally, we present an analytical expression to quantify the effective spatial DoF in the near-field, revealing that they are also constrained by the EBRD. Notably, the maximum spatial DoF is achieved with a ULA configuration, outperforming a square URA in this regard.

The evolution of next-generation wireless networks demands intelligent, adaptive, and explainable decision-making for robust communication in dynamic environments. This paper presentsEnwar 2.0, the first agentic large language model (LLM) framework integrating adaptive retrieval-augmented generation (RAG) and chain-of-thought (CoT) reasoning into situation-aware and explainable wireless network management.Enwar 2.0introduces two specialized agents: a transformer-fusion (TransFusion)-based beam prediction agent and an environment perception agent, both of which fuse multi-modal sensory inputs—including camera, LiDAR, radar, and GPS—from the DeepSense6G dataset. The beam prediction agent enables infrastructure-to-vehicle (I2V) target-in-the-loop beam tracking and real-time adaptation based on dynamic environmental conditions. In contrast, the environment perception agent provides situation-aware reasoning and justifications for beam decisions. Unlike its predecessor,Enwar 1.0, which relied on static knowledge bases (KBs) and text-only LLMs,Enwar 2.0is designed for CoT reasoning, leverages LLaMa3.2-3B/LLaMa3.1-8B/LLaMa3.3-70B for text-generation, the multi-modal capabilities of LLaMa 3.2, and employs LlamaIndex for fine-grained, dynamic context retrieval, eliminating retrieval ambiguities and enhancing response relevance. Numerical results show that the beam prediction agent achieves up to 90.0% Top-3 accuracy at$t+3$, effectively predicting optimal beam selections three time steps ahead. Overall,Enwar 2.0achieves state-of-the-art performance, with up to 89.7%/83.5% interpretation/perception correctness, 81.6%/80.9% faithfulness, and 89.9%/88.2% relevancy. In comparison, the baseline pretrained LLaMa3 models without adaptive RAG achieves up to 80.3%/77.3% correctness, and the baseline without RAG performs significantly worse at 67.1%/64.8%. Additionally,Enwar 2.0reduces processing time by over 100% relative to the baseline, while its adaptive RAG improves performance by up to 13.7% compared to static RAG.

Next generation wireless networks focus on improving spectral efficiency (SE) while reducing power consumption and hardware cost. Reconfigurable intelligent surfaces (RISs) offer a viable solution to meet these requirements. In order to enhance the SE, index modulation (IM) has been regarded as one of the enabling technologies via the transmission of additional information bits over the transmission media such as subcarriers, antennas and spatial paths. In this work, we explore the usage of spatial paths and introduce spatial path IM (SPIM) for RIS-aided massive multiple-input multiple-output (mMIMO) systems. Thus, the proposed framework improves the network efficiency and the coverage with the use of RIS while SPIM provides SE improvement. In order to perform SPIM, we exploit the spatial diversity of the millimeter wave channel and assign the index bits to the spatial patterns of the channel between the base station and the users through RIS. We introduce a low complexity approach for the design of hybrid beamformers, which are constructed by the steering vectors corresponding to the selected spatial path indices for SPIM-mMIMO. Furthermore, we conduct a theoretical analysis on the SE of the proposed SPIM approach, and derive the SE relationship between the SPIM-based hybrid beamforming and fully digital (FD) beamforming. Via numerical simulations, we validate our theoretical results and show that the proposed SPIM approach presents an improved SE performance, even higher than that of the use of FD beamformers while using a few RF chains.

In line with the AI-native 6G vision, explainability and robustness are crucial for building trust and ensuring reliable performance in millimeter-wave (mmWave) systems. Efficient beam alignment is essential for initial access, but deep learning (DL) solutions face challenges, including high data collection overhead, hardware constraints, lack of explainability, and susceptibility to adversarial attacks. This paper proposes a robust and explainable DL-based beam alignment engine (BAE) for mmWave multiple-input multiple-output (MIMO) systems. The BAE uses received signal strength indicator (RSSI) measurements from wide beams to predict the best narrow beam, reducing the overhead of exhaustive beam sweeping. To overcome the challenge of real-world data collection, this work leverages a site-specific digital twin (DT) to generate synthetic channel data closely resembling real-world environments. A model refinement via transfer learning is proposed to fine-tune the pre-trained model residing in the DT with minimal real-world data, effectively bridging mismatches between the digital replica and real-world environments. To reduce beam training overhead and enhance transparency, the framework uses deep Shapley additive explanations (SHAP) to rank input features by importance, prioritizing key spatial directions and minimizing beam sweeping. It also incorporates the Deep k-nearest neighbors (DkNN) algorithm, providing a credibility metric for detecting out-of-distribution inputs and ensuring robust, transparent decision-making. Experimental results show that the proposed framework reduces real-world data needs by 70%, beam training overhead by 62%, and improves outlier detection robustness by up to 8.5×, achieving near-optimal spectral efficiency and transparent decision making compared to traditional softmax based DL models.

Ultra-massive multiple-input multiple-output (UM-MIMO) technology is a key enabler for 6G networks, offering exceptional high data rates in millimeter-wave (mmWave) and Terahertz (THz) frequency bands. The deployment of large antenna arrays at high frequencies transitions wireless communication into the radiative near-field, where precise beam alignment becomes essential for accurate channel estimation. Unlike far-field systems, which rely on angular domain only, near-field necessitates beam search across both angle and distance dimensions, leading to substantially higher training overhead. To address this challenge, we propose a discrete Fourier transform (DFT) based beam alignment to mitigate the training overhead. We highlight that the reduced path loss at shorter distances can compensate for the beamforming losses typically associated with using far-field codebooks in near-field scenarios. Additionally, far-field beamforming in the near-field exhibits angular spread, with its width determined by the user's range and angle. Leveraging this relationship, we develop a correlation interferometry (CI) algorithm, termed CI-DFT, to efficiently estimate user angle and range parameters. Simulation results demonstrate that the proposed scheme achieves performance close to exhaustive search in terms of achievable rate while significantly reducing the training overhead by 87.5%.

Integrated artificial intelligence (AI) and communication has been recognized as a key pillar of 6 G and beyond networks. In line with AI-native 6 G vision, explainability and robustness in AI-driven systems are critical for establishing trust and ensuring reliable performance in diverse and evolving environments. This paper addresses these challenges by developing a robust and explainable deep learning (DL)-based beam alignment engine (BAE) for millimeter-wave (mmWave) multiple-input multiple-output (MIMO) systems. The proposed convolutional neural network (CNN)-based BAE utilizes received signal strength indicator (RSSI) measurements over a set of wide beams to accurately predict the best narrow beam for each UE, significantly reducing the overhead associated with exhaustive codebook-based narrow beam sweeping for initial access (IA) and data transmission. To ensure transparency and resilience, the Deep k-Nearest Neighbors (DkNN) algorithm is employed to assess the internal representations of the network via nearest neighbor approach, providing human-interpretable explanations and confidence metrics for detecting out-of-distribution inputs. Experimental results demonstrate that the proposed DL-based BAE exhibits robustness to measurement noise, reduces beam training overhead by 75 % compared to the exhaustive search while maintaining near-optimal performance in terms of spectral efficiency. Moreover, the proposed framework improves outlier detection robustness by up to$5 \times$and offers clearer insights into beam prediction decisions compared to traditional softmax-based classifiers.

The next generation wireless networks focus on improving the energy and spectral efficiency (SE/EE) of the communication systems in response to the demand for massive number of users and data rate. In this work, we aim to achieve the enhancement of SE and EE by employing index modulation (IM) techniques for reconfigurable intelligent surface (RIS)-aided communication systems. While RIS offers the network efficiency and improve the coverage, IM provides SE improvement by the transmission of additional index bits. In IM, we utilize the indices of the spatial paths between the base station and the user through the RIS. We introduce a low complexity approach for the design of hybrid beamformers, which are constructed by the steering vectors corresponding to the selected spatial path indices for IM. Via numerical experiments, we show that the proposed approach presents an improved SE performance, even higher than that of the use of fully-digital beamformers while using a few RF chains.

With the deployment of large antenna arrays at high-frequency bands, future wireless communication systems are likely to operate in the radiative near-field. Unlike far-field beam steering, near-field beams can be focused within a spatial region of finite depth, enabling spatial multiplexing in both the angular and range dimensions. This paper derives the beamdepth for a generalized uniform rectangular array (URA) and investigates how array geometry influences the near-field beamdepth and the limits where near-field beamfocusing is achievable. To characterize the near-field boundary in terms of beamfocusing and spatial multiplexing gains, we define the effective beamfocusing Rayleigh distance (EBRD) for a generalized URA. Our analysis reveals that while a square URA achieves the narrowest beamdepth, the EBRD is maximized for a wide or tall URA. However, despite its narrow beamdepth, a square URA may experience a reduction in multiuser sum rate due to its severely constrained EBRD. Simulation results confirm that a wide or tall URA achieves a sum rate of 3.5× more than that of a square URA, benefiting from the extended EBRD and improved spatial multiplexing capabilities.

Reconfigurable intelligent surfaces (RISs) are gaining significant attention as a key enabler of future wireless networks. However, its practical deployment is hindered by challenges in modeling and integration. Existing analytical approaches often depend on idealized assumptions, limiting their ability to reflect the complexity of real-world environments. In this work, we explore the integration of digital twin (DT) technology with ray tracing (RT), enabling a more accurate representation of practical scenarios and bridging the gap between theoretical models and the real-world implementation of RIS. RT allows accurate prediction of signal behavior, such as received signal strength indicator (RSSI) levels, without the need for expensive experimental measurements. We evaluate the performance of the DT with RT in three RIS-aided setups: a single RIS-aided, cascaded RISs-aided, and RIS partitioning. Our results show that the proposed DT model closely matches the experimental RSSI data with an error margin below 1 dB for single and partitioned RIS setups and under 2 dB for cascaded RISs system. These findings highlight the potential of DT with RT simulations as a practical tool for performance evaluation and optimization of RIS-assisted wireless systems.

This paper proposes an innovative approach for enhancing network capacity and coverage by integrating cell-free massive multiple-input multiple-output (CF-mMIMO) networks with reconfigurable intelligent surfaces (RISs). A significant challenge in leveraging RIS-assisted CF-mMIMO lies in the cooperative beam training across multiple access points (APs) and RISs, complicated by the passive nature of reflective elements and the complexity channel state information (CSI) acquisition in millimeter wave mMIMO systems. To address these challenges, we develop a multi-agent deep reinforcement learning (MA-DRL) framework that jointly designs beamforming and reflection codebooks for distributed APs and RISs, eliminating the need for CSI and relying solely on received power measurements feedback. The joint beamforming and reflection codebook design problem is decomposed into two sub-problems: one for beam codebook design at APs and another for sequential reflection codebook design at RISs. We employ transfer learning to speed up learning convergence and reduce computational complexity for training multiple RISs. Additionally, we introduce an AP and RIS selection scheme that improves overall energy efficiency and reduces backhaul overhead. Extensive simulations demonstrate that our proposed MA-DRL approach curtails number of beams significantly, thereby outperforming the widely adopted discrete Fourier transform (DFT) codebooks by achieving an 84% reduction in beam training overhead. Our findings suggest that increasing the number of passive RISs allows putting more APs into idle mode, leading to substantial savings in hardware and energy costs.

Artificial intelligence (AI) is expected to significantly enhance radio resource management (RRM) in sixth-generation (6G) networks. However, the lack of explainability in complex deep learning (DL) models poses a challenge for practical implementation. This paper proposes a novel explainable AI (XAI)-based framework for feature selection and model complexity reduction in a model-agnostic manner. Applied to a multi-agent deep reinforcement learning (MADRL) setting, our approach addresses the joint sub-band assignment and power allocation problem in cellular vehicle-to-everything (V2X) communications. We propose a novel two-stage systematic explainability framework leveraging feature relevance-oriented XAI to simplify the DRL agents. While the former stage generates a state feature importance ranking of the trained models using Shapley additive explanations (SHAP)-based importance scores, the latter stage exploits these importance-based rankings to simplify the state space of the agents by removing the least important features from the model’s input. Simulation results demonstrate that the XAI-assisted methodology achieves ~97% of the original MADRL sum-rate performance while reducing optimal state features by ~28%, average training time by ~11%, and trainable weight parameters by ~46% in a network with eight vehicular pairs.

Ultra massive multiple-input-multiple-output (UM-MIMO) technology has emerged as a promising candidate for 6G networks, offering ultra-high spectral efficiency in wireless systems. The transition to large antenna arrays specially at high-frequency bands is fundamentally transforming wireless communication from the traditional far-field to the near-field realm. This transition poses a distinct challenge in channel estimation due to the associated pilot overhead from large antenna arrays and the absence of angular sparsity in near-field spherical wavefronts. However, polar-domain sparsity remains achievable, advocating the use of polar codebooks over traditional angle-based ones. Nevertheless, the size of the polar codebook presents a significant challenge, necessitating sampling of distance and angle points across the entire near-field. In this work, we investigate the near-field channel estimation techniques while identifying the boundaries of polar domain sparsity with minimal pilot overhead. We propose a novel polar codebook, which leverages our findings from sparsity analysis and exploits the beam-focusing properties of the near-field. Unlike existing work, the proposed polar codebook design is agnostic to user range information and has considerably reduced dimensions. Capitalizing on this new polar codebook, we introduce the beam focused simultaneous orthogonal matching pursuit (BF-SOMP) algorithm for efficient near-field channel estimation. To further improve the channel estimation accuracy, we then present a refinement procedure that iterates over off-grid angle and range samples to enhance the estimation accuracy. Simulation results demonstrate that the proposed polar codebook based algorithms outperform contemporary methods in terms of improved normalized mean square error (NMSE) and reduced computational complexity. When compared to existing channel estimation methods, the proposed algorithms achieve an NMSE improvement of 6 − 7 dB at low and high SNR values with 32 pilots, while utilizing a codebook nearly half the size of existing ones.

This paper presents an end-to-end (E2E) autoencoder learning framework that relies on unsupervised deep learning for the joint design of millimeter wave (mmWave) probing beams and hybrid precoding matrices in multi-user communication systems. Our model utilizes prior channel observations to achieve two main objectives: designing a compact set of probing beams and predicting off-grid radio frequency (RF) beamforming vectors. The E2E learning framework optimizes probing beams in an unsupervised manner, concentrating sensing power on promising spatial directions based on the environment. To this aim, we develop a neural network architecture respecting RF chain constraints and model received signal strength (RSS) using complex-valued convolutional layers. The autoencoder is trained to directly produce RF beamforming vectors for hybrid architectures based on projected RSS indicators (RSSIs). Once RF beamforming vectors for multi-users are predicted, baseband digital precoders are designed by accounting for multi-user interference. The autoencoder neural network is trained E2E in an unsupervised manner with a customized loss function aimed at maximizing RSS. In a system with 64 antennas, 4 RF chains, and 4 users, our approach requires only 8 probing beams to design RF beamforming vectors, compared to the conventional predefined codebooks with 64 or 128 beams.

Beam-squint is a challenging issue in ultra-wideband systems, e.g., terahertz (THz) integrated sensing and communications (ISAC). In order to compensate for the loss due to beam-squint, this paper leverages index modulation in spatial domain, which enables the transmission of additional information bits to improve the spectral efficiency (SE). Specifically, a joint antenna and spatial path index modulation (JASPIM) technique is proposed by exploiting the spatial diversity of both antenna and path indices. We present a hybrid beamforming technique with JASPIM for ISAC, wherein the analog beamformers are designed in accordance with the radar targets and the communications user. Numerical simulations demonstrate that our JASPIM-ISAC approach exhibits a significant SE improvement even higher than that of the use of fully digital beamformers in the presence of beam-squint.

The integration of reconfigurable intelligent surfaces (RIS) and artificial noise (AN) significantly enhances physical layer security (PLS) in wireless networks, provided that RIS’s phase shifts are precisely optimized to prevent security vulnerabilities. This paper introduces a reinforcement learning (RL)based algorithm designed to optimize the phase shifts in RIS-partitioning-aided PLS systems operating in the millimeter wave (mm-Wave), without requiring channel state information (CSI) for any users. The RL algorithm optimizes the phase shifts by efficiently selecting the best beam from a predefined codebook for different partitions, which simultaneously enhances the intended signal for legitimate users and increases the effectiveness of AN on eavesdroppers, thereby maximizing the system’s secrecy capacity (SC) and addressing the inherent non-convex challenges. Additionally, the paper details the development of an experimental testbed that provides essential data to refine the algorithm. The numerical results from the testbed highlight the significant impact of RIS partitioning in PLS, which can enhance the SC by an average of 55% over the full RIS scenario, and confirm the effectiveness of the RL-based algorithm in reducing computational complexity by approximately 80% compared to the exhaustive search algorithm.

In terahertz (THz) wideband systems, beam-squint causes deviations in the generated beam directions at different subcarriers due to the use of subcarrier-independent analog beamformers. In order to combat the performance loss due to beam-squint effect, this work employs spatial path index modulation (SPIM) to improve the spectral efficiency (SE) performance of the overall system, thereby compensating the loss due to beam-squint. Specifically, SPIM allows the transmission of additional information bits to the receiver via modulating the indices of the spatial paths. The proposed approach is evaluated in a THz integrated sensing and communications (THz-ISAC) scenario, wherein the beamformer design allows generating multiple beams toward both radar targets and the communications user. Numerical simulations demonstrate that the proposed approach exhibits significant SE performance even higher than that of the use of fully digital beamformers without SPIM in the presence of beam-squint.

Ultra massive multiple-input-multiple-output (UM-MIMO) has emerged as a prospective capacity-enhancing technology for next generation (NG) networks. To harness its full potential, accurate channel estimation with low pilot overhead becomes paramount. The combination of large antenna arrays and high frequency bands transitions the wireless communication from the far-field to the near-field realm. This transition presents a unique challenge, as channel sparsity in the angular domain becomes unattainable in the presence of near-field spherical wavefronts. Nonetheless, polar-domain sparsity is achievable, allowing existing near-field channel estimation methods to use polar codebooks over the classical angle-based codebooks. However, a major challenge in utilizing polar domain sparsity is the size of the polar codebook, demanding sampling of both distance and angle points across the entire near-field. In this work, we investigate near-field channel response in the beam space domain to redefine the boundaries of polar domain sparsity. We propose a novel polar codebook leveraging our results of sparsity analysis and beamfocusing property of near-field. Unlike existing work, the proposed polar codebook design is agnostic to user range information and has considerably reduced dimensions. Exploiting the new polar codebook, we present the beam focused simultaneous orthogonal matching pursuit (BF-SOMP) algorithm for efficient near-field channel estimation. Simulation results demonstrate that the proposed algorithm surpasses contemporary methods in terms of normalized mean square error (NMSE) and reduced computational complexity.

Reconfigurable intelligent surfaces (RISs) play a vital role in future wireless systems with the capability of enhancing propagation environments by intelligently reflecting the signals toward the target receivers. However, optimal tuning of the phase shifters at the RIS is challenging due to the passive nature of reflective elements and the high complexity of acquiring channel state information (CSI). Furthermore, the joint active beamforming and RIS reflection beam design is a tedious task due to the high computational complexity and the dynamic nature of the wireless environment. Today’s cellular networks establish data transmission by relying on pre-defined generic beamforming codebooks, which are neither site-specific nor adaptive to the changes in the wireless environment. Moreover, identifying the best beam is typically performed using an exhaustive search approach that prohibits the use of large codebook sizes due to the resulting high beam training overhead. Depending merely on the binary received signal strength, this work develops a multi-agent deep reinforcement learning (MA-DRL) framework that jointly designs the active and the passive reflection beam codebooks for the BS and the RIS, reflectively. To accelerate learning convergence and reduce the search space, the proposed model divides the RIS into multiple partitions and associates beam patterns to the surrounding environments with low computational complexity. Moreover, a hierarchical beam training solution is proposed to further reduce the beam training overhead of the single-beam training approach. Simulation results show that the proposed MA-DRL approach can provide a 97% beam training overhead reduction over the discrete Fourier transform (DFT) codebook.

As the demand for wireless connectivity continues to soar, the fifth generation and beyond wireless networks are exploring new ways to efficiently utilize the wireless spectrum and reduce hardware costs. One such approach is the integration of sensing and communications (ISAC) paradigms to jointly access the spectrum. Recent ISAC studies have focused on upper millimeter-wave and low terahertz bands to exploit ultrawide bandwidths. At these frequencies, hybrid beamformers that employ fewer radio-frequency chains are employed to offset expensive hardware but at the cost of lower multiplexing gains. Wideband hybrid beamforming also suffers from the beam-split effect arising from the subcarrier-independent (SI) analog beamformers. To overcome these limitations, we introduce a spatial path index modulation (SPIM) ISAC architecture, which transmits additional information bits via modulating the spatial paths between the base station and communications users. We design the SPIM-ISAC beamformers by estimating both radar and communications parameters through our proposed beam-split-aware algorithms. We then develop a family of hybrid beamforming techniques – hybrid, SI, subcarrier-dependent analog-only, and beam-split-aware beamformers – for SPIM-ISAC. Numerical experiments demonstrate that the proposed approach exhibits significantly improved spectral efficiency performance in the presence of beam-split when compared with even fully digital non-SPIM beamformers.

Reconfigurable intelligent surfaces (RISs) are envisioned to play a pivotal role in future wireless systems with the capability of enhancing propagation environments by intelligently reflecting the signals toward the target receivers. However, the optimal tuning of the phase shifters at the RIS is a challenging task due to the passive nature of reflective elements and the high complexity of acquiring channel state information (CSI). Conventionally, wireless systems rely on pre-defined reflection beamforming codebooks for both initial access and data transmission. However, these existing pre-defined codebooks are commonly not adaptive to the environments. Moreover, identifying the best beam is typically performed using an exhaustive search that leads to high beam training overhead. To address these issues, this paper develops a multi-agent deep reinforcement learning framework that learns how to jointly optimize the active beamforming from the BS and the RIS-reflection beam codebook relying only on the received power measurements. To accelerate learning convergence and reduce the search space, the proposed model divides the RIS into multiple partitions and associates beam patterns to the surrounding environments with low computational complexity. Simulation results show that the proposed learning framework can learn optimized active BS beamforming and RIS reflection codebook. For instance, the proposed MA-DRL approach with only 6 beams outperforms a 256-beam discrete Fourier transform (DFT) codebook with a 97% beam training overhead reduction.

Cell-free massive MIMO (CF-mMIMO) is a transformative wireless network technology that surmounts conventional cellular network limitations concerning coverage, capacity, and interference management. Despite offering numerous benefits, CF-mMIMO also presents significant challenges, particularly in signal processing and power allocation. This paper introduces an unsupervised learning framework for downlink (DL) power allocation in CF-mMIMO networks, utilizing only large scaling fading coefficients instead of the hard-to-obtain exact user equipment (UE) locations or channel state information. We consider the sum spectral efficiency (sum-SE) optimization objective and investigate two distinct precoding schemes-maximum ratio (MR) and regularized zero-forcing (RZF)-for multi-antenna access points (APs). A custom loss function is formulated to maximize the sum-SE at each UE while accounting for pilot contamination and ensuring that power budget constraints are satisfied at each AP. The proposed unsupervised learning approach circumvents the arduous task of training data computations typically required in supervised learning methods, bypassing the use of conventional complex optimization methods and heuristic methodologies. The simulation results demonstrate that the proposed unsupervised learning approach outperforms existing methods in terms of SE, showcasing an improvement up to 20%. The proposed unsupervised neural network also approximates the optimal solutions generated by convex solvers while significantly reducing computational complexity.

This paper introduces a reconfigurable intelligent surface (RIS)-assisted grant-free non-orthogonal multiple access (GF-NOMA) scheme. To ensure the power reception disparity required by the power domain NOMA (PD-NOMA), we propose a joint user clustering and RIS assignment/alignment approach that maximizes the network sum rate by judiciously pairing user equipments (UEs) with distinct channel gains, assigning RISs to proper clusters, and aligning RIS phase shifts to the cluster members yielding the highest cluster sum rate. Once UEs are acknowledged with the cluster index, they are allowed to access their resource blocks (RBs) at any time requiring neither further grant acquisitions from the base station (BS) nor power control as all UEs are requested to transmit at the same power. In this way, the proposed approach performs an implicit over-the-air power control with minimal control signaling between the BS and UEs, which has shown to deliver up to 20% higher network sum rate than benchmark GF-NOMA and grant-based optimal (OPT) PD-NOMA schemes depending on the network parameters. The given numerical results also investigate the impact of UE density, RIS deployment, and RIS hardware specifications on the overall performance of the proposed RIS-aided GF-NOMA scheme.

Reconfigurable intelligent surface (RIS) assisted wireless systems require accurate channel state information (CSI) to control wireless channels and improve both the bandwidth and energy efficiency. However, CSI acquisition is non-trivial for two reasons: 1) the passive nature of RIS does not allow transceiving and processing pilot signals, and 2) the dimensions of the cascaded channel between transceivers increases with the large number of RIS elements, which yields high training overhead and computational complexity. While prior art has mainly focused on frequency-flat channel estimation, this paper proposes novel data-driven and compressive sensing based approaches for estimating both frequency-flat and frequency-selective cascaded channels of RIS-assisted multi-user millimeter-wave large multiple input multiple output (MIMO) systems with limited training overhead. The proposed methods exploit the common sparsity property among the different subcarriers and the double-structured sparsity property of the angular cascaded channel matrices as different angular cascaded channels observed by different users share completely common non-zero rows and user-specific column supports. The proposed data-driven cascaded channel estimation approaches use denoising neural networks to accurately detect channel supports. Alternatively, when data-training capabilities are not available, the compressive sensing based orthogonal matching pursuit (OMP) approach relies on sparsity properties and applies simultaneous OMP to detect the channel supports. Simulation results show that the pilot overhead required by the proposed scheme is lower than existing schemes. When compared to other OMP approaches that achieve an NMSE gap of 5 to 6 dB with respect to the Oracle least square lower bound, the proposed algorithms reduce the lower bound gap to only 1 dB, while reducing complexity by more than two orders of magnitude.

Reconfigurable intelligent surface (RIS) assisted wireless systems require accurate channel state information (CSI) to control wireless channels and improve overall network performance. However, CSI acquisition is non-trivial due to the passive nature of RIS, and the dimensions of the cascaded channel between transceivers increase with the large number of RIS elements, which requires high training overhead. Prior art has considered frequency-selective channel estimation without considering the beam squint effect in wideband systems, severely degrading channel estimation performance. This paper proposes a novel data-driven approach for estimating wideband cascaded channels of RIS-assisted multi-user millimeter-wave massive multiple-input multiple-output (MIMO) systems with limited training overhead, explicitly considering the effect of beam squint. To circumvent the beam squint effect, the proposed method exploits the common sparsity property among the different subcarriers as well as the double-structured sparsity property of the users’ angular cascaded channel matrices. The proposed data-driven cascaded channel estimation approach exploits denoising neural networks to detect channel supports accurately. Compared to beam squint effect agnostic traditional orthogonal matching pursuit (OMP) approaches, the proposed data-driven approach achieves 5-6dB less normalized mean square error (NMSE) and reduces the lower bound gap to only 1dB for the oracle least-square benchmark.

Millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) systems typically employ hybrid mixed signal processing to avoid expensive hardware and high training overheads. However, the lack of fully digital beamforming at mmWave bands imposes additional challenges in channel estimation. Prior art on hybrid architectures has mainly focused on greedy optimization algorithms to estimate frequency-flat narrowband mmWave channels, despite the fact that in practice, the large bandwidth associated with mmWave channels results in frequency-selective channels. In this paper, we consider a frequency-selective wideband mmWave system and propose two deep learning (DL) compressive sensing (CS) based algorithms for channel estimation. The proposed algorithms learn critical apriori information from training data to provide highly accurate channel estimates with low training overhead. In the first approach, a DL-CS based algorithm simultaneously estimates the channel supports in the frequency domain, which are then used for channel reconstruction. The second approach exploits the estimated supports to apply a low-complexity multi-resolution fine-tuning method to further enhance the estimation performance. Simulation results demonstrate that the proposed DL-based schemes significantly outperform conventional orthogonal matching pursuit (OMP) techniques in terms of the normalized mean-squared error (NMSE), computational complexity, and spectral efficiency, particularly in the low signal-to-noise ratio regime. When compared to OMP approaches that achieve an NMSE gap of$\mathrm {\{4-10\}\,\,dB}$with respect to the Cramer Rao Lower Bound (CRLB), the proposed algorithms reduce the CRLB gap to only$\mathrm {\{1-1.5\}\,\,dB}$, while reducing complexity by two orders of magnitude.

Cell-free massive MIMO communications is an emerging network technology for 5G wireless communications wherein distributed multi-antenna access points (APs) serve many users simultaneously. Most prior work on cell-free massive MIMO systems assume time-division duplexing mode, although frequency-division duplexing (FDD) systems dominate current wireless standards. The key challenges in FDD massive MIMO systems are channel-state information (CSI) acquisition and feedback overhead. To address these challenges, we exploit the so-called angle reciprocity of multipath components in the uplink and downlink, so that the required CSI acquisition overhead scales only with the number of served users, and not the number of AP antennas nor APs. We propose a low complexity multipath component estimation technique and present linear angle-of-arrival (AoA)-based beamforming/combining schemes for FDD-based cell-free massive MIMO systems. We analyze the performance of these schemes by deriving closed-form expressions for the mean-square-error of the estimated multipath components, as well as expressions for the uplink and downlink spectral efficiency. Using semi-definite programming, we solve a max-min power allocation problem that maximizes the minimum user rate under per-user power constraints. Furthermore, we present a user-centric (UC) AP selection scheme in which each user chooses a subset of APs to improve the overall energy efficiency of the system. Simulation results demonstrate that the proposed multipath component estimation technique outperforms conventional subspace-based and gradient-descent based techniques. We also show that the proposed beamforming and combining techniques along with the proposed power control scheme substantially enhance the spectral and energy efficiencies with an adequate number of antennas at the APs.

Device-to-Device (D2D) communication is a viable network technology that can potentially enhance the spectral and energy efficiency of cellular networks. To exploit this benefit in D2D-underlaid cellular networks, the co-channel interference between D2D and cellular users should be properly managed. In this paper, we propose a joint channel allocation (CA) and power control (PC) scheme to mitigate interference in a D2D underlaid cellular system modeled as a random network using stochastic geometry. The novel aspect of the proposed CA scheme is that it enables D2D links to share resources with multiple cellular users as opposed to one as previously considered in the literature. The PC scheme compensates for large-scale path-loss effects by employing distance-dependent path-loss parameters with an estimation error margin. Closed-form expressions for the coverage probability of cellular links, D2D links, and the sum rate of the D2D links are derived in terms of the allocated power, density of the D2D links, and the path-loss exponent. Simulation results demonstrate an enhancement of 10%-40% for the cellular and D2D coverage probabilities, and 35% for spectral efficiency.

Device-to-Device (D2D) communications underlaying cellular networks is a viable network technology that can potentially increase spectral utilization and improve power efficiency for proximity-based wireless applications and services. However, a major challenge in such deployment scenarios is the interference caused by D2D links when sharing the same resources with cellular users. In this paper, we propose a channel allocation (CA) scheme together with a set of three power control (PC) schemes to mitigate interference in a D2D underlaid cellular system modeled as a random network using the mathematical tool of stochastic geometry. The novel aspect of the proposed CA scheme is that it enables D2D links to share resources with multiple cellular users as opposed to one as previously considered in the literature. Moreover, the accompanying distributed PC schemes further manage interference during link establishment and maintenance. The first two PC schemes compensate for large-scale path-loss effects and maximize the D2D sum rate by employing distance-dependent path-loss parameters of the D2D link and the base station, including an error estimation margin. The third scheme is an adaptive PC scheme based on a variable target signal-to-interference-plus-noise ratio, which limits the interference caused by D2D users and provides sufficient coverage probability for cellular users. Closed-form expressions for the coverage probability of cellular links, D2D links, and sum rate of D2D links are derived in terms of the allocated power, density of D2D links, and path-loss exponent. The impact of these key system parameters on network performance is analyzed and compared with previous work. Simulation results demonstrate an enhancement in cellular and D2D coverage probabilities, and an increase in spectral and power efficiency.

Device-to-Device (D2D) communication is a promising technology that can potentially enhance the spectral and energy efficiency of cellular networks. To exploit this benefit in D2D-underlaid cellular networks, the co-channel interference between D2D and cellular users should be properly managed. In this paper, we propose a distributed power control scheme to mitigate interference in a D2D underlaid cellular system modeled as a random network using the mathematical tool of stochastic geometry. The proposed PC scheme compensates for large-scale path-loss effects by employing distance-dependent path-loss parameters of the D2D link and the base station, including an estimation error margin. Closed-form expressions for the coverage probability of cellular links, D2D links, and the sum rate of D2D links are derived in terms of the allocated power, density of D2D links, and path-loss exponent. The coverage performance of both cellular and D2D users is analyzed, and the analytical results are validated through simulations. Experimental results demonstrate the efficacy and advantages of our proposed scheme over other schemes by an enhancement of 20%-30% for the cellular and D2D coverage probabilities, and an increase in spectral efficiency by 60%.