Compact ultra-massive antenna arrays (CUMA) share key characteristics with holographic communication systems, featuring densely spaced and individually controlled antenna elements that enable precise manipulation of electromagnetic waves. In this paper, we investigate the spectral efficiency (SE) of CUMA deployed within some constrained physical space. Departing from prior works that assume ideal isotropic antennas, we derive a closed-form expression for the SE assuming a line-of-sight (LoS) channel at the electromagnetic level, explicitly accounting for mutual coupling and antenna orientation. The analysis reveals that the channel gain is highly sensitive to both the array orientation and individual antenna directions. In the single-user case, our results show that the optimal orientation of the user array is either aligned parallel or perpendicular to the signal direction, depending on the inter-element spacing. Notably, near-optimal channel gain is achieved when individual antennas are oriented perpendicular to the signal direction. In the multi-user case, we further optimize transceiver configurations under mutual coupling constraints. Simulation results confirm that SE is strongly influenced by the directional alignment of user antennas and array placement in the near-field regime. CUMA significantly outperforms traditional half-wavelength spaced arrays in terms of SE when constrained to the same physical aperture.

This paper investigates a new integrated sensing and communication (ISAC) scheme for underwater acoustic (UWA) networks based on deep reinforcement learning, referred to as Deep UWA-ISAC (DeepUSC). Specifically, we consider a UWA-ISAC system, where an autonomous underwater vehicle (AUV) transmits the collected environmental data to the buoy, while sensing the sea area to monitor the unauthorized mobile target. The expected communication rate over a given navigation period is maximized by jointly optimizing the AUV's beamforming and trajectory, subject to the constraints on the average signal-to-noise ratio requirement for target sensing as well as the navigation mission, collision avoidance, and maximum transmit power limit of the AUV. Three key challenges for DeepUSC are: (i) long propagation delays in the UWA-ISAC system may cause interference from the previous echo to the current ISAC signal; (ii) the mobility pattern of the target is unknown in advance; and (iii) the AUV navigation-oriented ISAC problem is a long-term optimization problem as the navigation mission typically lasts for a long period. To circumvent the above challenges, DeepUSC is developed based on a specific partially observable Markov decision process model termed episode task, where each navigation period is considered as an episode and the navigation mission corresponds to the episode task. Through judicious design of a reward function and action selection policy, DeepUSC can satisfy various preset constraints without requiring prior knowledge of the target's mobility. Besides, to enable efficient learning in episode tasks, we propose an episodic experience replay mechanism that dynamically prioritizes high-value recent experiences and utilizes all experiences generated within each episode to jointly train the neural network. Simulation results demonstrate that compared with benchmarks, DeepUSC yields a higher communication rate while satisfying all constraints, converges faster, and is more robust against different simulation setups.

Consider an integrated sensing and communication (ISAC) system where a base station (BS) employs a full-duplex radio to simultaneously serve multiple users and detect a target. The detection performance of the BS may be compromised by self-interference (SI) leakage. This paper investigates the feasibility of SI cancellation (SIC) through the application of symbol-level precoding (SLP). We first derive the target detection probability in the presence of the SI. We then formulate an SLP-based SIC problem, which optimizes the target detection probability while satisfying the quality of service requirements of all users. The formulated problem is a nonconvex fractional programming (FP) problem with a large number of equality and inequality constraints. We propose a penalty-based block coordinate descent (BCD) algorithm for solving the formulated problem, which allows for efficient closed-form updates of each block of variables at each iteration. Finally, numerical simulation results are presented to showcase the enhanced detection performance of the proposed SIC approach.

Cell-free multiple-input-multiple-output generally uses a star topology for superior communication but faces high costs due to long cables. An economical alternative, the stripe topology, is suitable for specific deployments but cannot meet user demands in densely populated areas due to limited fronthaul capacity. To address these limitations, we propose a hybrid network structure combining stripe and star topologies, ensuring system performance while reducing deployment costs. In such a network, joint precoding and fronthaul compression is considered to maximize system sum-rate and an alternating optimization (AO) algorithm is proposed. However, the AO algorithm involves an iterative process and complex matrix calculations, making it unsuitable for practical applications. To deal with this issue, we propose a low-complexity iterative gradient descent (IGD) algorithm with simple matrix operations. To further reduce online computational complexity, we propose a novel deep unfolding neural network (DUNN) scheme, which is interpretable and scalable, based on the IGD algorithm. Simulation results show that the hybrid topology significantly improves system capacity compared to the stripe-only topology. Additionally, the DUNN achieves a tradeoff between the achievable sum-rate performance and the corresponding computational complexity.

To reduce the computational complexity of near-field channel estimation for extremely large-scale multiple-input multiple-output (XL-MIMO) systems, the concept ofvirtual cell subarraysin subarray hybrid precoding architectures, is firstly given. Multiple subarrays with strong correlation can be flexibly and dynamically combined, to jointly extract relevant features of the channels. Based on this, an undersampling matching and oversampling refinement pursuit (UMORP) algorithm is proposed, which can detect the channel parameters of cell subarrays through an undersampling dictionary constructed by analog phase shifters. This approach facilitates the estimation of the channel of a single cell subarray with much fewer pilots and lower hardware capability requirement. Then, a multiple-path decoupled spatial extrapolation (MPDSE) scheme is proposed for fully dimensional channel reconstruction, which orthogonally decouples multiple paths from the received signal of a single cell subarray firstly, and then utilize the spatial correlation between adjacent cell subarrays to extrapolate the channels with low cost. Moreover, to further reduce the computational complexity in XL-MIMO systems, a spatial multiple cell subarrays joint extrapolation (SMCJE) scheme is also proposed. Based on the spatial geometry among cell subarrays, only several cell suabrrays are utilized to jointly estimate the distances and reconstruct the fully dimensional near-field channel, achieving a low level of computational complexity. Our simulation results verify that the proposed schemes perform better in XL-MIMO systems, while requiring fewer pilots and exhibiting much lower computational complexity.

Deploying an extremely large-scale reconfigurable intelligent surface (XL-RIS) can significantly improve the performance of a RIS-assisted communication system. However, the array aperture and deployment of the XL-RIS affects the radiated field region in which the base station (BS) and the user are located, which in turn affects the performance improvement. In this paper, we have jointly optimized a deployment scheme and phase-shift matrix in XL-RIS-aided communication system, aiming to maximize the user’s received signal-to-noise ratio (SNR). Firstly, we incorporate the far-field and near-field channel into a unified far- or near-field (FoN) model to simplify the SNR analysis and optimization on RIS deployments. Secondly, based on the FoN approach, we derive an expression for the user’s received SNR and formulate an optimization problem to jointly optimize the RIS deployment and phase-shift matrix in order to maximize the user’s received SNR. Thirdly, we summarize the relationship between the RIS array aperture and deployment and the radiated field region in which the BS and the user are located, and propose an optimized closed-form solution for the RIS deployment and phase-shift matrix. Finally, we validate the effectiveness of the proposed scheme through simulation results.

Autonomous aerial vehicles (AAVs) with cache are considered as an efficient technique to enhance serving capabilities of traditional wireless networks in terms of network coverage and capacity. However, with the introduction of AAVs, new challenges such as trajectory design and AAV-user association occur. In this paper, we consider a caching AAV-assisted wireless network and formulate a user fairness problem by jointly optimizing AAV-user association, trajectory design, and bandwidth allocation of the AAVs, which is mixed-integer and non-convex. In order to find solutions, we decompose the original problem into three subproblems and propose an iterative algorithm based on block alternating descent and successive convex approximation methods. In addition, computational complexity is analyzed. Finally, simulation results validate the efficiency of the proposed algorithm, compared to benchmark algorithms.

More recently, the cell-free massive multiple-input multiple-output (CF -mMIMO) has gained much attention as it can significantly reduce the path loss and improve the user experience. However, the application of CF-mMIMO is limited by the payload and delays caused by channel state information (CSI) exchange among access points (APs). To solve this problem, we use the distributed learning (DL) framework to design a fully local receiver. We set up a distributed deep neural network (DNN) model at each AP that takes the local estimated CSI as input, labeled by the centralized minimum mean square error (MMSE) receiver. As no CSI exchange is needed, the AP can produce a powerful local receiver without any payload and delays on the fronthaul during the whole process. Simulation results have shown that the DL-aided receiver can significantly improve the spectral efficiency over traditional local receivers. Moreover, as we take the large-scale fading into account and apply appropriate hyperparameters to accelerate the model convergence, our model can converge fast and be available for both fixed and moving users.

Visibility region (VR) is known as a key channel characteristic appeared in extra-large massive MIMO (XL-MIMO) systems, which can be exploited to facilitate low-complexity transmission design. Existing VR recognition method requires an a priori location-Vrdataset, with which a user's VR can be estimated given its location. This dataset is constructed by selecting some beacon users (BUs) to estimate the VR at their locations via uplink training. Constrained by the available training resource and possible environmental variation, practical size of the dataset is usually limited; how to efficiently select BUs for better VR recognition accuracy is therefore important. To this end, we propose and compare three BU selection methods, including random selection, minimum spacing constrained (MSC) selection, and a more sophisticated method (denoted as dynamic boundary refining, DBR) which utilizes partial of BUs for exploring unknown environment, while selecting the other BUs for further refining already-estimated VR region boundaries. Simulation results show that with a small number of BUs, both MSC and DBR achieve similar VR recognition performance and outperform random selection; as the number of BUs becomes larger, DBR achieves the best recognition accuracy.

By decoupling the dedicated radio frequency (RF) chain into transmit RF (TX RF) chain and receive RF (RX RF) chain, the asymmetrical system can flexibly equip the downlink/uplink array with different number of TX/RX RF chain according to the practical demand in a massive multiple-input multiple-output Internet of Things (IoT) network. To reduce cost and power consumption, this paper maximizes the uplink resource efficiency (RE) under Weichselberger channel model by designing transmit covariance matrices and receive antenna selection (RAS). In IoT networks with multiple IoT nodes, we propose an alternate optimization algorithm to iteratively optimize transmit covariance matrices and RAS by exploiting statistical channel state information. Specifically, for correlated channels, we propose a penalty method-based algorithm for RAS which utilizes Dinkelbach’s transform and linear relaxation to tackle the intractable fractional function and binary constrain, respectively. Compared with greedy search, the proposed algorithm has lower complexity without much loss of performance. For independent identically distributed channels, we simplify the RE maximization problem and provide the necessary conditions of the optimal number of receive antennas and transmit power. Finally, the validness of our conclusions as well as the effectiveness of proposed algorithms are illustrated by numerical simulations.

In “Channel Estimation for Massive MIMO-OTFS System in Asymmetrical Architecture,” by Chen et al., a two-stage channel estimation scheme is proposed based on the input-output relationship of orthogonal time frequency space (OTFS) modulation in asymmetrical architecture. This correspondence provides some comments and corrections to the derivation of the OTFS input-output relationship published in [1].

To glean the benefits offered by massive multi-input multi-output (MIMO) systems, channel state information must be accurately acquired. Despite the high accuracy, the computational complexity of classical linear minimum mean squared error (MMSE) estimator becomes prohibitively high in the context of massive MIMO, while the other low-complexity methods degrade the estimation accuracy seriously. In this paper, we develop a novel rank-1 subspace channel estimator to approximate the maximum likelihood (ML) estimator, which outperforms the linear MMSE estimator, but incurs a surprisingly low computational complexity. Our method first acquires the highly accurate angle-of-arrival (AoA) information via a constructed space-embedding matrix and the rank-1 subspace method. Then, it adopts thepost-receptionbeamforming to acquire the unbiased estimate of channel gains. Furthermore, a fast method is designed to implement our new estimator. Theoretical analysis shows that the extra gain achieved by our method over the linear MMSE estimator grows according to the rule of O(log10M), while its computational complexity islinearlyscalable to the number of antennasM. Numerical simulations also validate the theoretical results. Our new method substantially extends the accuracy-complexity region and constitutes a promising channel estimation solution to the emerging massive MIMO communications.

In this paper, an efficient near-field channel estimation algorithm, and a novel cascade channel reconstruction scheme, with significantly reduced pilot overhead and computational complexity, are proposed for hybrid extra large-scale reconfigurable intelligent surface (XL-RIS)-aided multi-input multi-output (MIMO) systems. A unique hybrid XL-RIS architecture is devised, in which the elements at the designed central subarray, and many specially selected discrete elements, are active, while others are passive. Meanwhile, a damped Newtonized orthogonal matching pursuit algorithm combining the planar and spherical wave models (DNOMP-CPSW) is proposed, in which the angle and distance parameters of multipaths are estimated through the received signals of the central subarray and the discrete active elements respectively, and the near-field channel can be reconstructed accurately with low pilot overhead and computational complexity. Moreover, to decrease the cost of cascade channel reconstruction in the considered system, a separate channel estimation scheme based on the decoupling operation (SCEDO) is proposed, which estimates the two separate channels with only 3 pilots and reduced computational complexity, and then reconstruct the cascade channel. Furthermore, the phase shift strategy of the XL-RIS with 2-bits quantization is devised, which can increase the energy of the received signal and maintain the set order to estimate multipaths in different stages of the SCEDO scheme, to improve the accuracy of the estimates, and enhance the reliability of the separate channel estimation. Simulation results verify that the proposed DNOMP-CPSW algorithm and the hybrid XL-RIS phase shift strategy can enhance the performance of the considered system. Compared with other schemes, the SCEDO scheme can reconstruct the cascade channel efficiently, with much reduced pilot overhead and computational complexity.

To harness the benefits of both reconfigurable intelligent surface (RIS) and dynamic metasurface antenna (DMA), we consider the hybrid RIS and DMA assisted multiple-input multiple-output (MIMO) multiple-access channel (MAC) over spatially correlated Rician fading, in which multiple multi-antenna users send the transmitted signals to the DMA-based base station (BS) with the assistance of a RIS. The objective is to maximize the achievable ergodic sum-rate by jointly designing the transmit covariance matrix of users, the phase shift matrix of RIS, and the DMA weight matrix at BS only with statistical channel state information. By capitalizing on large random matrix theory, a closed-form asymptotic ergodic sum-rate is first obtained. Then, we propose a modified water-filling algorithm to design the optimal transmit covariance matrix under the power consumption and specific absorption rate constraints. Next, we design the phase shift matrix of RIS via the projected gradient ascent algorithm, subject to the non-convex unit-modular constraint. To find the constrained DMA weight matrix, we further resort to the optimal solution of the unconstrained DMA problem and adopt the alternating optimization method. The proposed algorithm is numerically shown to improve the sum-rate compared to the baseline schemes, verifying the effectiveness of the proposed schemes.

Extra large-scale multiple-input multiple-output (XL-MIMO) is a key technology for future wireless communication systems. This paper considers the effects of visibility region (VR) at the base station (BS) in a non-stationary multi-user XL-MIMO scenario, where only partial antennas can receive users’ signal. In time division duplexing (TDD) mode, we first estimate the VR at the BS by detecting the energy of the received signal during uplink training phase. The probabilities of two detection errors are derived and the uplink channel on the detected VR is estimated. In downlink data transmission, to avoid cumbersome Monte-Carlo trials, we derive a deterministic approximate expression for ergodic average energy efficiency (EE) with the regularized zero-forcing (RZF) precoding. In frequency division duplexing (FDD) mode, the VR is estimated in uplink training and then the channel information of detected VR is acquired from the feedback channel. In downlink data transmission, the approximation of ergodic average EE is also derived with the RZF precoding. Invoking approximate results, we propose an alternate optimization algorithm to design the detection threshold and the pilot length in both TDD and FDD modes. The numerical results reveal the impacts of VR estimation error on ergodic average EE and demonstrate the effectiveness of our proposed algorithm.

Ultra-reliable and low-latency communication (URLLC) is recognized as the most challenging use case for the next generation of wireless networks. Existing research on URLLC is based on the regular pilot (RP) scheme, which is tough to ensure a high transmission rate with stringent latency and reliability requirements due to the impact of finite blocklength, especially in massive connectivity scenarios. In this paper, we propose to use generalized superimposed pilot (GSP) scheme for URLLC transmission in massive multi-input multi-output (mMIMO) systems. Distinguishing from the conventional superimposed pilot (SP) scheme, the GSP scheme eliminates mutual interference between the pilot and data, where the data length is optimized, and the data symbols are precoded to spread over the whole transmission block. With the GSP scheme, we first formulate a weighted sum rate maximization problem by jointly optimizing the data length, pilot power, and data power and then derive closed-form results, including suboptimal data length and achievable rate lower bounds with maximum-ratio combining (MRC) and zero-forcing (ZF) detectors, respectively. Based on the closed-form results, we provide the corresponding iterative algorithms for the MRC and ZF cases where the problems are transformed into geometry program format by using log-function and successive convex approximation methods. Finally, the performance of the RP, SP, and GSP schemes are compared through simulation results, which reflect the superiority and robustness of the GSP scheme in URLLC scenarios.

A reconfigurable intelligent surface (RIS) with active elements is an augmented version of an RIS. By equipping all or part of RIS elements with signal processing capabilities, the channel estimation and the design of RIS phases can be further extended, yielding an improvement in the spectral efficiency (SE). In this paper, we first present a novel sensing RIS structure which is efficient for hardware implementation. Unlike partial active elements in previous structures, all elements are available to the RF chains via switches, which enables the traditional channel estimation methods and channel extrapolation to be implemented. Moreover, we make a comprehensive analysis and comparison with other RIS structures from the perspective of channel state information (CSI) acquisition. Considering the large-scale of RIS and base station (BS) array, we model the channel between the user and the RIS, the RIS and the BS using a near-field channel model. Based on the structured channel model, we propose a low-overhead channel reconstruction protocol through a parameter-extracting method, while the training overhead and complexity are also analyzed. In addition, we investigate the RIS elements’ activation strategy to further reduce the training overhead. Finally, numerical results demonstrate that the proposed scheme achieves accurate channel estimation with low overhead, which can also enhance the achievable SE.

We investigate a new deployment form of reflective intelligent surface (RIS), which aims at enhancing the quality of service of a main communication system in a target region, while without the need of changing its transmission protocol and scheme (i.e., the RIS is “transparent” to the main system). To this end, we mathematically formulate a coverage enhancement problem, where a RIS is used transparently in the sense that the BS can be unaware of its existence, while the minimum channel link strength, measured from every BS antenna to any point in the target region, can be maximized. The formulated problem is non-convex with mixed discrete-continuous variables. To tackle this challenge, we recast it into a convex feasibility problem via spatial sampling and semi-definite relaxation. Based on a derived analytical upper bound on the link strength difference between any two location points, we further characterize the coverage-similarity region of a given location, and accordingly propose an improved spatial sampling scheme for efficient implementation. Simulation results show that the proposed transparent RIS design achieves better coverage performance than benchmark schemes. More importantly, it can effectively improve the communication performance without affecting the transmission scheme originally adopted by the main communication system.

The application of distributed reconfigurable intelligent surface (RIS) to massive access multiple-input single-output (MISO) is significant to extend the communication coverage. In this paper, a novel framework is proposed in distributed RIS-aided massive access MISO systems with supporting non-orthogonal multiple access (NOMA) and orthogonal multiple access (OMA) transmissions simultaneously, where a joint active and passive beamforming scheme is designed to fully eliminate the inter-cluster interference and improve the channel gains of prioritized users, respectively. Based on the proposed framework, firstly, we derive two exact channel statistics to characterize the equivalent channel gains of prioritized and non-prioritized users, respectively. Then, by taking into account the influence of imperfect channel state information (CSI) and successive interference cancellation (SIC), the approximate expressions of outage probability and ergodic rate for all users of one cluster under MISO-NOMA and MISO-OMA transmissions are analyzed to obtain their corresponding system throughput. Moreover, by utilizing the above results, we also determine the diversity order and high slope of these users to attain more viewpoints. Finally, simulation results prove our analyses and reveal that: 1) enhancing the estimated accuracy of CSI and the ability of SIC process can remarkably enhance the system performance; 2) the performance of priority users will be significantly improved with the increase of the number of reflecting elements and Rician factor; 3) heterogeneous quality of service requirements and deployment behaviors of users are beneficial for NOMA, while homogenous settings are competitive for OMA.

In this paper, we propose a low-complexity mode switching design in a simultaneously transmitting and reflecting reconfigurable intelligent surface (STAR-RIS) aided multi-antenna non-orthogonal multiple access (NOMA) system via exploiting channel angle information only. Particularly, a location-based assignment algorithm is firstly provided to perform NOMA pairing, then the equal gain transmission at base station and cophase matching criterion at STAR-RIS are adopted to improve the performance of reflected user and the effective channel gain of paired users, respectively. On the basis of to the above design, we study the exact channel statistics in two different cases to further derive the outage probability of reflected and transmitted users. Numerical results are presented to demonstrate our analyses and reveal that: 1) the various system factors have markedly impacts on our proposed design; and 2) the low-complexity design can reduce the overheads of channel estimation and signal processing at the expense of extremely low performance loss.

In this paper, we investigate a distributed reconfigurable intelligent surface aided massive access in multipleinput single-output non-orthogonal multiple access system with imperfect channel state information (CSI) and successive interference cancellation (SIC). In particular, a novel active and passive beamforming scheme are designed to fully eliminate the intercluster interference and improve the effective channel gain of the prioritized users, respectively. To study the performance of the proposed scheme, the exact channel statistics are derived to further analyze the outage probability of each user within a cluster. Finally, simulation results are presented to prove our theoretical analyses and reveal that: 1) enhancing the estimated accuracy of CSI and the ability of SIC process can significantly enhance the outage performance; 2) the proposed zero-forcing based scheme can obtain a higher system throughput compared to previous designs.

The orthogonal time frequency space (OTFS) is poised to become a pivotal technology for the next generation of mobile communications, due to its inherent robustness against Doppler shift. By combining OTFS technology with massive multiple-input multiple-output (MIMO) technology, users can experience high-quality communication services even in highly mobile scenarios. In this letter, we extend the massive MIMO-OTFS system to an asymmetrical architecture with unequal number of transceiver radio frequency chains. To overcome the channel inconsistency and recover the downlink channel by partial uplink channel, we utilize coprime patterns and propose a channel estimation algorithm that firstly extracts the angle parameters from the virtual array and then estimates the remaining channel parameters, which effectively reduces the three-dimensional search space to two dimensions. Our numerical simulations demonstrate that the proposed algorithm enhances the accuracy of channel estimation with much lower complexity.

Driven by the development of time-sensitive applications, short packet transmission (SPT) design has become the key point in the ultra-reliable and low-latency communications (URLLC) area. The primary challenge in it is that the delay caused by pilot overhead cannot be neglected. To deal with this issue, this paper presents a frame structure adopting partial-superimposed-pilot (PSP) scheme for SPT. The key of PSP scheme is that the number of data symbols is equal to the available blocklength, and the number of pilot symbols transmitted with data in the training stage needs to be optimized. Under the finite blocklength regime, we first derive a closed-form lower bound achievable rate of an uplink massive MIMO system with imperfect pilot removal for maximal-ratio-combining (MRC) receiver. Then, we formulate a weighted sum rate maximization problem by jointly optimizing the pilot length, pilot power, and data power. We derive a closed-form solution of optimal pilot length. Using the log-function and successive convex approximation (SCA) method, we develop an iterative optimization framework to find a locally optimal power solution. For comparison, the conventional frame structures based on complete-superimposed-pilot (CSP) and regular pilot (RP) schemes are also shown. Simulation results indicate that the proposed PSP scheme is superior to the existing CSP and RP schemes.

With the rapid development of 5G, massive machine-type communication is expected to experience significant growth, leading to severe random access collisions. To address this issue, we first adopt deep neural networks to detect random access collisions by learning the features of the received signals. Based on the collision-detection results, we propose a double-contention random access (DCRA) scheme, with which the base station can schedule one more contention process for devices experiencing collisions. To fully harness the collision-resolution capability of the proposed DCRA scheme, we further analyze its performance and illustrate how to tune the backoff parameters to optimize the network throughput. It is revealed that the maximum throughput of the DCRA scheme depends on the number of random access preambles and the collision recognition accuracy. The corresponding optimal backoff parameters are then obtained, which greatly facilitates implementations in practice. Simulation results show that with a high collision recognition accuracy, the proposed scheme can achieve significant throughput improvement.

In this paper, we consider a downlink multiple-input single-output non-orthogonal multiple access (MISO-NOMA) network with several legitimate users and a eavesdropper using successive interference cancellation (SIC). The purpose of this paper is to maximize the secrecy performance of the MISO-NOMA network by designing the transmit power between the legitimate users and the artificial jamming. Explicitly, the secrecy sum rate of the MISO-NOMA network is to be maximized by optimizing the transmit beamforming vectors and the artificial jamming vector, subject to the required quality of service of each legitimate user, the artificial jamming beamforming design constraint and the SIC decoding condition. Due to the non-convexity of the optimization problem, we reformulate the original problem into an equivalent optimization problem and then provide a successive convex approximation based iterative algorithm for solving it. Simulation results demonstrate that the proposed optimization scheme outperforms the existing schemes.

In this paper, we study a distributed reconfigurable intelligent surface-aided downlink multiple-input multiple-output non-orthogonal multiple access systems with random distributed users, where the signal alignment can be achieved within the paired users by controlling discrete phase shifts. In particular, a unified precoder and decoder are provided to cancel inter-cluster interference. To evaluate the performance of proposed framework, the near and far users’ channel statistics with Nakagami-m fading are derived. Further, the approximate expressions of average outage probability are obtained by utilizing the proposed channel statistics, and the corresponding diversity orders can also be obtained to acquire more insights. Finally, simulation results reveal that not only discrete phase shifts design by selecting a 3bit resolution can realize the near optimal performance but also the diversity gain can be significantly improved by increasing the number of reflection elements.

Energy harvesting (EH)-enabled cellular Internet of Things (IoT) is a promising solution to handle the charging and accessing of massive IoT nodes. However, limited by the high-frequency band of future 5G, the radius of the small base station (SBS) is reduced, hence greatly increasing the cost of the network operators (NOs). In this article, we consider the joint cell association, cell sleeping (CS), and incentive decision problem for EH-aided cellular IoT with full-duplex (FD) self-backhauls. We formulate a Stackelberg game to investigate the coordination between the utilities of NO and energy transmitters (ETs), where both the features of FD self-backhauls and CS are introduced to reduce the expense of NO. We then propose an alternative direction algorithm to solve the equilibrium of the game efficiently, where the relationship of the formulated constraints and variables are utilized to transform the original problem into two subproblems. We propose a two-level Lagrangian relaxation to solve the first subproblem, while the other is proved to be convex and solved by an efficient iteration. Simulation results demonstrate the benefits of our algorithm in utility improvement and expense reduction. Moveover, it shows that our algorithm can obtain high efficiency by adjusting the tradeoff between the number of active SBS and transmitting power of ETs according to the network deployment.

The combination of reconfigurable intelligent surface (RIS) and non-orthogonal multiple access (NOMA) has been recognized as a critical method to improve the sixth generation networks performance. In this paper, a distributed RIS-aided downlink multiple-input multiple-output (MIMO) NOMA systems with discrete phase shifts are studied, where the channel directions from base station to paired users can be manipulated with the assistance of RISs by employing the concept of signal alignment. In order to ensure base station can flexibly serve some users with NOMA and others users with spatial division multiple access, a unified precoder and decoder are provide to cancel inter-cluster interference. Subsequently, the channel statistics over Nakagami-$m$fading channels are derived for near and far users. In particular, considering the cascade channel gain may exists two different cases, Beaulieu series is further adopted to characterize their corresponding cumulative distribution function. In what follows, the outage probability and ergodic rate for three situations within one cluster are derived by utilizing the obtained channel statistics, respectively. Based on the derived results, we also analyze the diversity order and high signal-to-noise ratio slope to provide essential insights into the considered systems. Finally, simulation results are presented to reveal that: 1) selecting the setting of 3-bits resolution can realize a near-aligned performance for our proposed systems; 2) the cascade channel statistic caused by RIS can be evaluated with any number of RIS element and any channel gain via Beaulieu series.

Multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM), a fundamental transmission scheme, promises high throughput and robustness against multipath fading. However, these benefits rely on the efficient detection strategy at the receiver and come at the expense of the extra bandwidth consumed by the cyclic prefix (CP). We use the iterative orthogonal approximate message passing (OAMP) algorithm in this paper as the prototype of the detector because of its remarkable potential for interference suppression. However, OAMP is computationally expensive for the matrix inversion per iteration. We replace the matrix inversion with the conjugate gradient (CG) method to reduce the complexity of OAMP. We further unfold the CG-based OAMP algorithm into a network and tune the critical parameters through deep learning (DL) to enhance detection performance. Simulation results and complexity analysis show that the proposed scheme has significant gain over other iterative detection methods and exhibits comparable performance to the state-of-the-art DL-based detector at a reduced computational cost. Furthermore, we design a highly efficient CP-free MIMO-OFDM receiver architecture to remove the CP overhead. This architecture first eliminates the intersymbol interference by buffering the previously recovered data and then detects the signal using the proposed detector. Numerical experiments demonstrate that the designed receiver offers a higher spectral efficiency than traditional receivers. Finally, over-the-air tests verify the effectiveness and robustness of the proposed scheme in realistic environments.

Federated learning (FL) has become a promising solution to train a shared model without exchanging local training samples. However, in the traditional cloud-based FL framework, clients suffer from limited energy budget and generate excessive communication overhead on the backbone network. These drawbacks motivate us to propose an energy-aware hierarchical federated learning framework in which the edge servers assist the cloud server to migrate the local models from the clients. Then a joint local computing power control and client association problem is formulated in order to minimize the training loss and the training latency simultaneously under the long-term energy constraints. To solve the problem, we recast it based on the general Lyapunov optimization framework with the instantaneous energy budget. We then propose a heuristic algorithm, which takes the importance of local updates into account, to achieve a suboptimal solution in polynomial time. Numerical results demonstrate that the proposed algorithm can reduce the training latency compared to the scheme with greedy client association and myopic energy control, and improve the learning performance compared to the scheme in which the associated clients transmit their local models with the maximal power.

Reconfigurable intelligent surface (RIS) stands out as a promising technology by enhancing the electromagnetic wave propagation environment with its passive reflecting elements. In this paper, we focus on the ergodic sum-rate analysis and maximization of the RIS-assisted uplink multiuser multiple-input multiple-output (MIMO) multiple-access channel (MAC) under Rician fading by exploiting full statistical channel state information (CSI). The spatial correlations at the base station, the users, and the RIS are also considered. By using the replica method originated in statistical physics, the closed-form asymptotic ergodic sum-rate of the system is first derived in the large-system regime on the account of the unique channel structure of the RIS-assisted MIMO-MAC system. Then, based on the derived asymptotic ergodic sum-rate, we propose an alternating optimization (AO) algorithm to jointly design the transmit covariance matrix of users and the phase-shifting matrix of RIS with full statistical CSI. Simulation results are also demonstrated to verify the accuracy of the derived asymptotic ergodic sum-rate and the superiority of the proposed AO algorithm. The results reveal that the derived closed-form asymptotic ergodic sum-rate matches very well with the Monte Carlo results even for a small number of antennas, and the proposed AO algorithm can achieve up to 10 bps/Hz gain at high signal-to-noise (SNR) regime, which is therefore valuable for future RIS-assisted MIMO-MAC system designs.

In application scenarios such as Internet of Things, a large number of energy receivers (ERs) exist and line-of-sight (LOS) propagation could be common. Considering this, we investigate wireless energy transfer (WET) in extra-large massive MIMO Rician channels. We derive analytical expressions of the received net energy for different schemes, including 1) training-based WET, where the ER sends beacon signal for channel training and the energy transmitter (ET) uses the channel estimate for energy beamforming, 2) LOS beamforming, where the ET transmits to the LOS direction of the ER, and 3) energy harvesting, which allows an ER to harvest the training energy from the other ERs. We derive a path loss threshold for switching between training and LOS beamforming-based WET. We further show that the WET scheme selection of one ER is not affected by the other ERs, and the energy harvested from training is minimal in practice. With these insights, we propose an algorithm for the multi-ER scenario, which minimizes the power consumption by iteratively updating the WET scheme selection and power allocation for all ERs. Simulations show that the proposed algorithm achieves near-optimal performance as compared to exhaustive searching, while with much lower implementation complexity.

Reconfigurable intelligent surface (RIS) is an emerging technology to enhance wireless communication in terms of energy cost and system performance by equipping a considerable quantity of nearly passive reflecting elements. This study focuses on a downlink RIS-assisted multiple-input multiple-output (MIMO) wireless communication system that comprises three communication links of Rician channel, including base station (BS) to RIS, RIS to user, and BS to user. The objective is to design an optimal transmit covariance matrix at BS and diagonal phase-shifting matrix at RIS to maximize the achievable ergodic rate by exploiting the statistical channel state information at BS. Therefore, a large-system approximation of the achievable ergodic rate is derived using the replica method in large dimension random matrix theory. This large-system approximation enables the identification of asymptotic-optimal transmit covariance and diagonal phase-shifting matrices using an alternating optimization algorithm. Simulation results show that the large-system results are consistent with the achievable ergodic rate calculated by Monte-Carlo averaging. The results verify that the proposed algorithm can significantly enhance the RIS-assisted MIMO system performance.

This study considers a downlink power consumption problem for unmanned aerial vehicles (UAVs)‐assisted wireless communications, in which UAVs are used as aerial base stations to provide service for the ground users and equipped with a directional antenna of fixed beamwidth. Moreover, the on‐board circuit power of UAV is taking into consideration. The authors derive a closed‐form expression for the optimal flying altitude and number of UAVs by minimising the total power consumption under the users' rate requirements in the given coverage area. The numerical simulation and theoretical results show that the optimal flying altitude of UAVs depends on the beamwidth of the directional antenna at UAVs, the on‐board circuit power of UAVs, and the rate constraint of each user.

Environmental monitoring is one of the fundamental applications of the Internet of Things (IoT), and caching in energy harvesting-aided IoT is a promising solution to handle the energy charging of the IoT nodes in the vast monitoring area. However, the growth of the requirements for monitoring area and accuracy brings huge infrastructure costs to the network operators (OP), especially for the multiple OPs scenario. In this article, we utilize wireless virtualization to enable the IoT node sharing between multiple OPs in cache-enabled energy harvesting-aided IoT, so as to improve the utility of the OPs. A Stackelberg game is formulated to jointly handle the IoT node sharing and energy transmission incentives between OPs and energy transmitters. Then, the knapsack problem, convex and linear programming are utilized to approximate the game through problem transformation and derivations. On the basis of that, an alternative direction algorithm is proposed to solve the equilibrium efficiently. The simulation results verify the advantages of the proposed algorithm in utility improvement and fairness maintenance between multiple OPs.

Beamforming is an effective means to improve the quality of the received signals in multiuser multiple-input-single-output (MISO) systems. Traditionally, finding the optimal beamforming solution relies on iterative algorithms, which introduces high computational delay and is thus not suitable for real-time implementation. In this paper, we propose a deep learning framework for the optimization of downlink beamforming. In particular, the solution is obtained based on convolutional neural networks and exploitation of expert knowledge, such as the uplink-downlink duality and the known structure of optimal solutions. Using this framework, we construct three beamforming neural networks (BNNs) for three typical optimization problems, i.e., the signal-to-interference-plus-noise ratio (SINR) balancing problem, the power minimization problem, and the sum rate maximization problem. For the former two problems the BNNs adopt the supervised learning approach, while for the sum rate maximization problem a hybrid method of supervised and unsupervised learning is employed. Simulation results show that the BNNs can achieve near-optimal solutions to the SINR balancing and power minimization problems, and a performance close to that of the weighted minimum mean squared error algorithm for the sum rate maximization problem, while in all cases enjoy significantly reduced computational complexity. In summary, this work paves the way for fast realization of optimal beamforming in multiuser MISO systems.

The machine-to-machine (M2M) communication is an emerging technology that is widely utilized in a vast number of industrial Internet-of-Things (IIoT) applications. Due to the diversity of IIoT applications, provisioning of heterogeneous delay requirements of delay-sensitive machine type devices (MTDs) while optimizing the access efficiency of delay-tolerate MTDs becomes a critical challenge for M2M communications. To address this issue, a multigroup analytical framework for massive random access of M2M communications in IIoT is proposed in this article. Specifically, we consider delay-sensitive MTDs and delay-tolerate MTDs coexist in the network, and those MTDs are divided into multiple groups according to their delay requirements. The access behavior of each MTD is characterized by a double-queue model. Based on this model, the throughput and the mean access delay of each group are characterized. It is found that for each group, the mean access delay decreases as the throughput increases and is minimized when the throughput is maximized. To achieve the maximum throughput of delay-tolerate MTDs under delay constraints of delay-sensitive MTDs, the backoff parameters of delay-sensitive MTDs should be tuned according to the delay constraints while that of delay-tolerate MTDs should be tuned further according to the aggregate packet arrival rate and the number of MTDs in each group. It is further demonstrated that the optimal tuning of backoff parameters is robust against the burstiness of input traffic. The analysis sheds important light on the access design of M2M communications in IIoT with delay constraints.

Traffic delay is a key metric to measure the quality-of-service of next-generation wireless communication networks. In this paper, we consider a cloud radio access network architecture with a hierarchical structure of virtual controllers and multiple clusters of remote radio heads (RRHs). A high-level controller coordinates control plane decisions among local controllers and each local controller is in charge of a cluster of RRHs. Moreover, each local controller is equipped with one server for creating virtual machines (VMs) to execute the users' baseband processing tasks. Then, under the considered architecture, we aim to minimize the average delay consisting of task execution delay and signal transmission delay under total power constraint, by joint optimization of task scheduling and resource allocation, including VM allocation and RRH assignment. Due to the non-deterministic polynomial-time hardness (NP-hardness) of the joint optimization problem, we translate it into a matroid constrained submodular maximization problem and propose heuristic algorithms to find solutions with 0.5-approximation. Besides, both centralized and distributed control schemes are considered. In the centralized control scheme, all decisions about task scheduling, VM allocation, and RRH assignment are made in the high-level controller. But in the distributed control scheme, the high-level controller is only in charge of task scheduling based on graph theory and the local controllers are responsible for their respective VM allocation and RRH assignment. The simulation results show that the proposed algorithms can achieve better performance than the separate optimization of VM allocation and RRH assignment.

In this paper, we aim to minimize the network power consumption (NPC) in a downlink cloud radio access network. Not only the powers consumed at remote radio heads and fronthaul links for transmission, but also the power consumed at the baseband unit pool for computation is considered. We formulate a joint NPC minimization problem as a mixed timescale issue which can be regarded as a combination of two power minimization problems for computation and transmission, where the former is a slow timescale issue since task scheduling and computation resource allocation are usually executed in a large time space whereas the latter is a fast timescale issue due to the dependence on small-scale fading. To deal with timescale challenge, we introduce approximate results of the joint NPC minimization problem according to large system analysis and turn it into a slow timescale issue because the approximations are only dependent on statistical channel information. We propose an iterative coordinate descent algorithm based on branch-and-bound algorithm to find solutions to the joint NPC minimization problem. Numerical results show that the NPC decreases as the delay constraint increases but increases if the execution efficiency or computing capability of servers is degraded.

In this paper, we consider an uplink heterogeneous cloud radio access network (H-CRAN), where a macro base station (BS) coexists with many remote radio heads (RRHs). For cost savings, only the BS is connected to the baseband unit (BBU) pool via fiber links. The RRHs, however, are associated with the BBU pool through wireless fronthaul links, which share the spectrum resource with radio access networks. Due to the limited capacity of fronthaul, the compress-and-forward scheme is employed, such as point-to-point compression or Wyner-Ziv coding. Different decoding strategies are also considered. This paper aims to maximize the uplink ergodic sum-rate (SR) by jointly optimizing quantization noise matrix and bandwidth allocation between radio access networks and fronthaul links, which is a mixed time-scale issue. To reduce computational complexity and communication overhead, we introduce an approximation problem of the joint optimization problem based on large-dimensional random matrix theory, which is a slow time-scale issue, because it only depends on statistical channel information. Finally, an algorithm based on Dinkelbach's algorithm is proposed to find the optimal solution to the approximate problem. In summary, this paper provides an economic solution to the challenge of constrained fronthaul capacity and also provides a framework with less computational complexity to study how bandwidth allocation and fronthaul compression can affect the SR maximization problem.

In this paper, we consider the network power minimization problem in a downlink cloud radio access network (C-RAN), taking into account the power consumed at the baseband unit (BBU) for computation and the power consumed at the remote radio heads and fronthaul links for transmission. The power minimization problem for transmission is a fast time-scale issue, whereas the power minimization problem for computation is a slow time-scale issue. Therefore, the joint network power minimization problem is a mixed time-scale problem. To tackle the time-scale challenge, we introduce large system analysis to turn the original fast time-scale problem into a slow time-scale one that only depends on the statistical channel information. In addition, we propose a bound improving branch-and-bound algorithm and a combinational algorithm to find the optimal and suboptimal solutions to the power minimization problem for computation, respectively, and propose an iterative coordinate descent algorithm to find the solutions to the power minimization problem for transmission. Finally, a distributed algorithm based on hierarchical decomposition is proposed to solve the joint network power minimization problem. In summary, this paper provides a framework to investigate how execution efficiency and computing capability at BBU as well as delay constraint of tasks can affect the network power minimization problem in C-RANs.

This paper studies analog beam selection schemes of discrete Fourier transform (DFT) based hybrid beamforming systems. We first derive approximations of the achievable rates when maximum-ratio combining (MRC) receiver and maximum-ratio transmitting (MRT) precoder are used in the uplink and downlink, respectively. It is shown that the achievable rate of the hybrid beamforming system is improved with the increase of the number of radio frequency chains. Also, it is found that the orthogonality condition among the line-of-sight (LoS) paths from different users directly determines the interference cancellation capability of the MRC receiver or the MRT precoder. Based on our analytical results, we propose two novel DFT beam selection schemes, referred to as exhausted searching and per-user selection. Numerical results show that the first scheme achieves higher rate while the second one is a simple suboptimal strategy with low complexity, which is practically more attractive.

In this paper, we analyze the effect of phase noise on the downlink ergodic sum-rate of a cloud radio access network. The system comprises of one baseband processing unit (BBU) on the cloud server which coordinates M multi-antenna remote radio heads (RRHs) serving K single-antenna users using regularized zero-forcing precoding. We assume the BBU has all users' data and imperfect channel state information and communicate with RRHs via optical fibers which are referred to as fronthaul links. The effect of phase noise both at RRHs and users is also taken into consideration. A deterministic approximation of downlink ergodic sum-rate is derived based on large dimensional random matrix theory when the numbers of antennas at RRHs and users are asymptotically large with a fixed ratio. From simulation results, it is confirmed that the deterministic approximation is accurate and the effect of phase noise is shown to result in a significant reduction in system performance.

In this paper, we consider the uplink transmission of a heterogeneous cloud radio access network, where a macro base station (BS) and many remote radio heads (RRHs) coexist to serve user equipment units. For cost-savings, only the BS is connected to the baseband unit (BBU) pool via fiber links, whereas the RRHs are associated with the BBU pool through wireless fronthaul links with limited capacities. By employing Wyner-Ziv (WZ) coding scheme, the RRHs first compress the received signal and then transmit the corresponding quantized version to the BBU pool. We derive deterministic equivalent for ergodic uplink sum rate and use this result to jointly optimize quantization noise matrix and bandwidth allocation between radio access networks and fronthaul links. An algorithm based on Dinkelbach's algorithm is also proposed to determine the optimal solutions. Numerical results show that as the normalized fronthaul capacity increases, more bandwidth is allocated to radio access networks. Besides, uniform quantization with WZ coding across RRHs can achieve near-optimal performance under high signal-to-quantization-noise ratio.

Small-cell networks and massive multiple-input multiple-output (MIMO) systems are regarded as important candidate techniques for 5G communication systems. This paper considers a heterogeneous network composed of a macrocell tier overlaid with an extremely dense tier of small-cells. In the network, the macrocell base station (BS), which applies massive MIMO, does not only serve macro user equipment units but also provides wireless backhaul for small-cell access points (APs). The wireless backhaul shares the same spectrum resource with radio access networks without creating extra spectrum resources. However, due to the densification of small-cells, the inter- and intra-tier interferences become severe. To mitigate the interferences, we use the regularized zero-forcing precoding combined with a projection technique is used at the BS in downlink (DL) to avoid interference to the APs in uplink (UL). Meanwhile, the joint linear minimum mean square error detection is applied in UL to mitigate the inter-tier interference. We derive deterministic expressions for ergodic UL and DL sum rates (SRs) by leveraging the large-dimensional random matrix theory. The expressions only depend on statistical channel information and can be used to optimize the bandwidth division between radio access links and wireless backhaul, as well as the time allocation between DL and UL operation intervals. Numerical results show that the deterministic SR equivalents are accurate and that the proposed resource allocation method can significantly improve system performance.

In this paper, we consider a heterogeneous network in which a macro-cell tier is overlaid with a very dense tier of small cells. The macro-cell base station (BS) that applies a massive MIMO scheme not only serves the macro user equipment but also provides a wireless backhual for small-cell access points (APs). These APs serve their associated small-cell user equipment. A reverse time division duplex transmission protocol is utilized. To avoid interference toward the APs in the uplink (UL), regularized zero-forcing precoding combined with a projection technique is utilized at the BS in the downlink (DL). We derive deterministic expressions for ergodic UL and DL sum rates (SRs) under the assumption that perfect channel state information is available and use these results to optimize the spectrum division between radio access links and the wireless backhaul. Simulation results suggest that the deterministic SR approximations are accurate and that system performance can be significantly improved through the optimization of spectrum division.

In this paper we investigate the physical layer security for downlink MU-MIMO systems where the transmitter and the eavesdropper are equipped with multiple antennas, while the legitimate users have single antenna. This is a more challenging topic because the eavesdropper appears to be more powerful than the legitimate users. We propose a transmission scheme able to enhance secure transmission for legitimate users, by intentionally applying different beamforming matrices to pilot signals and data signals. The beamforming matrix for data signals is constructed in a way that legitimate users can derive the channel matrix experienced by data signals based on pilot signals, whereas the eavesdropper fails to obtain the channel matrix experienced by data signals even with the aid of pilot signals. Therefore coherent detection at the eavesdropper is impossible and the signal-plus-interference-to-noise (SINR) of the eavesdropper is significantly degraded. In addition, the proposed beamforming design is formulated as an optimization problem to maximize the minimum SINR. We theoretically prove that with this formulation, the resultant optimal SINR is not affected by the proposed format of beamforming matrix for data signals. Numerical results show that, the proposed beamforming design outperforms the conventional beamforming design in terms of ergodic secrecy sum rate. Furthermore, with suitable beamforming designs, the ergodic secrecy sum rate achieved will not be noticeably affected by the number of antennas at the eavesdropper, although it is significantly degraded with conventional beamforming design.

In this paper, downlink transmission schemes that are able to improve the energy efficiency of multi-cell massive distributed antenna systems (DAS) are investigated. We employ a power consumption model by considering the transmit power, the backhaul power, the uplink pilot transmit power and the circuit power, in contrast to existing works which focus on co-located antenna systems where the backhaul power is negligible. By applying random matrix theory and given the power consumption model, we derive the asymptotic energy efficiency achieved by two transmission schemes, first a full transmission scheme where all remote radio heads (RRHs) in a cell jointly transmit, and second single-RRH user association transmission scheme where each user is associated with only one RRH. A new algorithm to associate user with RRH is proposed based on the asymptotic energy efficiency. The proposed algorithm takes into consideration the power attached to antennas as well as the backhauling power of RRH. Simulations show that the proposed user association algorithm for DAS achieves higher energy efficiency than full transmission scheme and the baseline nearest RRH association scheme, especially when the number of antennas is large.

© 2015 IEEE. In this paper, we study the multiple-input multiple-output wiretap channel for simultaneous wireless information and power transfer, in which there is a base station (BS), an information-decoding (ID) user, and an energy-harvesting (EH) user. The messages intended to the ID user is required to be kept confidential to the EH user. Our objective is to design the optimal transmit covariance matrix at the BS for maximizing the ergodic secrecy rate subject to the harvested energy requirement for the EH user exploiting only statistical channel state information at the BS. To this end, we begin by deriving an approximation for the ergodic secrecy rate using large-dimensional random matrix theory and the method of Taylor series expansion. This approximation enables us to derive the asymptotic-optimal transmit covariance matrix that achieves the tradeoff for ergodic secrecy rate and harvested energy. The simulation results are provided to verify the accuracy of the approximation and show that a bigger rate-energy region can be achieved when the Rician factor increases or the path loss exponent decreases. We also show that when the transmit correlation increases or the distance between the eavesdropper and the BS decreases, the harvested energy will be increased, while the achieved ergodic secrecy rate decreases.

In this paper, we consider a cognitive radio (CR) network in which a secondary multiantenna base station (BS) attempts to communicate with multiple secondary users (SUs) using the radio frequency spectrum that is originally allocated to multiple primary users (PUs). Here, we employ partially-projected regularized zero-forcing (PP-RZF) precoding to control the amount of interference at the PUs and to minimize inter-SUs interference. The PP-RZF precoding partially projects the channels of the SUs into the null space of the channels from the secondary BS to the PUs. The regularization parameter and the projection control parameter are used to balance the transmissions to the PUs and the SUs. However, the search for the optimal parameters, which can maximize the ergodic sum-rate of the CR network, is a demanding process because it involves Monte-Carlo averaging. Then, we derive a deterministic expression for the ergodic sum-rate achieved by the PP-RZF precoding using recent advancements in large dimensional random matrix theory. The deterministic equivalent enables us to efficiently determine the two critical parameters in the PP-RZF precoding because no Monte-Carlo averaging is required. Several insights are also obtained through the analysis.

In this paper, a deterministic equivalent of ergodic sum rate and an algorithm for evaluating the capacity-achieving input covariance matrices for the uplink large-scale multiple-input multiple-output (MIMO) antenna channels are proposed. We consider a large-scale MIMO system consisting of multiple users and one base station with several distributed antenna sets. Each link between a user and an antenna set forms a two-sided spatially correlated MIMO channel with line-of-sight (LOS) components. Our derivations are based on novel techniques from large dimensional random matrix theory (RMT) under the assumption that the numbers of antennas at the terminals approach to infinity with a fixed ratio. The deterministic equivalent results (the deterministic equivalent of ergodic sum rate and the capacity-achieving input covariance matrices) are easy to compute and shown to be accurate for realistic system dimensions. In addition, they are shown to be invariant to several types of fading distribution.

In this paper, we analyze the ergodic sum-rate of a multi-cell downlink system with base station (BS) cooperation using regularized zero-forcing (RZF) precoding. Our model assumes that the channels between BSs and users have independent spatial correlations and imperfect channel state information at the transmitter (CSIT) is available. Our derivations are based on large dimensional random matrix theory (RMT) under the assumption that the numbers of antennas at the BS and users approach to infinity with some fixed ratios. In particular, a deterministic equivalent expression of the ergodic sum-rate is obtained and is instrumental in getting insight about the joint operations of BSs, which leads to an efficient method to find the asymptotic-optimal regularization parameter for the RZF. In another application, we use the deterministic channel rate to study the optimal feedback bit allocation among the BSs for maximizing the ergodic sum-rate, subject to a total number of feedback bits constraint. By inspecting the properties of the allocation, we further propose a scheme to greatly reduce the search space for optimization. Simulation results demonstrate that the ergodic sum-rates achievable by a subspace search provides comparable results to those by an exhaustive search under various typical settings.

In this paper, we consider the multi-cell downlink system with multiple base stations (BSs) and multiple single antenna users employing the BS cooperation. The channels between BSs and users have independent spatial correlations. All BSs jointly implement the regularized zero-forcing based on imperfect channel estimation. By utilizing the large dimensional random matrix theory, we first obtain the limiting distribution of the eigenvalues for a class of random Hermitian matrix. Based on this result, we derive a deterministic equivalent of ergodic sum rate for the multi-cell downlink system and obtain the optimal regularization parameter in the special case employing maximize the deterministic equivalent of ergodic sum rate. Numerical results show that the deterministic equivalent are accurate even for finite number of antenna and the channel estimation parameter almost do not affect the accuracy of the approximation.

In this paper, we focus on a general coordinated multi-point (CoMP) multiple input multiple-output (MIMO) system consisting of multiple users and multiple base stations (BSs) equipped with multiple antennas, respectively. An asymptotic ergodic mutual information expression and the capacity-achieving input covariance matrices for the system are derived employing novel techniques from large dimensional random matrix theory (RMT). The asymptotic regime is based on the assumption that the numbers of antennas at the transmitter and receiver approach to infinity with a fixed ratio. Our contributions are to extend the previous results to the general channel model with two-sided spatial correlation and line-of-sight (LOS), in which the transmit and receive correlation matrices are both generally nonnegative definite and the channel entries are non-Gaussian distributed. Simulations show that the asymptotic capacity is accurate even for finite number of antenna and invariant to all types of fading distribution.