Radio Access Network (RAN) disaggregation is emerging as a key trend in beyond 5G, as it offers new opportunities for more flexible deployments and intelligent network management. A relevant problem in disaggregated RAN is the functional split selection, which dynamically decides which baseband (BB) functions of a base station are kept close to the radio units and which ones are centralized. In this context, this paper firstly presents an architectural framework for supporting this concept relying on the O-RAN architecture. Then, the paper analyzes how the functional split can be optimized to adapt to the different load conditions while minimizing energy costs.
In the ever-evolving field of artificial neural networks and learning systems, complex-valued neural networks (CVNNs) have become a cornerstone, achieving exceptional performance in image processing and telecommunications. More precisely, in digital communication systems, CVNNs have been delivering significant results in tasks like equalization, channel estimation, beamforming, and decoding. Among the CVNN architectures, the complex-valued radial basis function neural network (C-RBF) stands out, especially when operating in noisy environments such as 5G multiple-input multiple-output (MIMO) systems. In such a context, this paper extends the classical shallow C-RBF to deep architectures, increasing its flexibility for a wider range of applications. Also, based on the parameter selection of the phase transmittance radial basis function (PT-RBF) neural network, we propose an initialization scheme for the deep C-RBF. Via rigorous simulations conforming to 3GPP TS 38 standards for digital communications, our method not only outperforms conventional initialization strategies like random, K-means, and constellation-based methods but it also seems to be the only approach to achieve successful convergence for deep C-RBF architectures. These findings pave the way to more robust and efficient neural network deployments in complex-valued digital communication systems.
The large number of individually controlled antennas in large-scale MIMO systems makes the complexity of uplink receive combining a serious challenge. This is why simple methods, such as maximum ratio combining, are often considered, even if lower spectral efficiency is attained. In this paper, we investigate the performance of a reduced-complexity combining scheme that exploits second-order statistics of the channel. Compared to maximum ratio combining, substantial improvements are achieved, especially when the number of active terminals increases. This is attained with limited computational cost and remarkable robustness to the imperfect channel statistics knowledge.
The low-resolution aware linear minimum mean-squared error (LRA-LMMSE) channel estimator for low-resolution MIMO receivers was shown to achieve lower mean-squared error (MSE) with the use of a comparator network, which is composed of several simple comparators with binary outputs. In this study, we improve the analysis for the MSE of a comparator network aided MIMO system employing 1-bit quantization by deriving a bound for the error when using an approximated version for the auto-correlation of the quantized signal. By simulations, we compare the obtained analytical expression with the state-of-the-art approximation, showing that the proposed analytical expression is closer to the actual MSE obtained numerically. We also simulate independent comparator networks, which appear to yield a lower bound on the MSE for specific configurations.
In this paper, constrained optimization methods with a sparsity-aware Newton-type direction update are applied to adaptive filtering. Different versions of the proposed algorithms, which include a sparsity-aware RLS and a CG algorithm, present lower computational complexity than traditional algorithms based on LMS-Newton, since the proposed algorithms require the use of a smaller Hessian estimate. The proposed algorithms are tested in acoustic echo cancellation and time-varying channel identification. Some of them present a computational reduction of 25% of that of the traditional LMS-Newton in the best case scenario. Most of the proposed algorithms present faster convergence than the benchmark algorithms in several scenarios.
In this paper, we have studied a wireless powered communication network with practically nonlinear energy harvesting (EH). An intelligent reflecting surface (IRS) has been deployed to assist both energy transfer and data transmissions. In particular, we assume there are two users with diverse communication types and the whole operation period is divided to two slots. In the first slot, one user behaves as downlink device requesting data from the access point (AP), while the other user acts as a rechargeable device collecting wireless energy from AP for harvesting its energy buffer. In the second slot, the rechargeable user will deploy harvested energy for supporting the data uploading to the AP. The IRS is applied for intelligently modifying the radio environment in both slots. Taking nonlinear EH model into account, we formulate a joint IRS beamforming and service slot optimization problem maximizing the uplink throughput with a given downlink throughput constraint. Subsequently, to deal with the design difficulty in nonlinear EH, we propose a variable substitution based problem reformulation and construct a series of convex approximations for iteratively approaching a suboptimal of the considered problem. Finally, via simulations, we validate the convergence of our proposed algorithm and illustrate the benefits of IRS beamforming in assisting wireless powered communication communication.
In this work, we analyze 2D-FFT FB as an alternative waveform to OTFS. By employing part of the OTFS codification, specifically part of the Symplectic Finite Fourier Transform (SFFT), the complex orthogonality of the system is guaranteed, resulting in a waveform that incorporates the main characteristics of OTFS systems such as robustness in high mobility scenarios and lower PAPR compared to OFDM. Furthermore, the use of filters well located in the time-frequency domain allows for better separation between channels, reducing out-of-band emissions. A comparison is done between both waveforms pointing out their similarities in the construction of the system as well as their benefits. Simulation results demonstrate a high compatibility between the techniques and a significant advantage of the 2D-FFT FB in terms of spectral efficiency and detection in the time-frequency domain.
In non-circular signals, the drawback of the fixed step-size of the widely linear complex-valued least mean square (WL-CNLMS) algorithm results in the algorithm being suboptimal. To address this problem, we introduce a combined step-size (CSS) strategy in the WL-CNLMS algorithm to improve the convergence performance. Also, to overcome the negative effects of non-Gaussian impulsive noise, we use a modified Huber (MH) function to improve the robustness of the algorithm. Finally, we propose a robust combined stepsize WL-CNLMS algorithm and apply it to beamforming. Simulation results show that the proposed algorithm not only enhances the convergence performance of the adaptive beamformer but also greatly improves its robustness.
In this paper, we analyze the performance of an algorithm for adaptive diffusion networks that controls the number of nodes sampled per iteration based on the estimation error. The goal of this solution is to keep the nodes sampled while the estimation error is high in magnitude, and to cease their sampling when it is sufficiently low. Our model shows that this approach can preserve the convergence rate in comparison with the case in which every node is sampled permanently, while slightly improving the steady-state performance.
This paper focuses on a large-scale wireless channel modeling for 3.5 GHz frequency, consisting of a fine-tuning of the Close-In (CI) and Floating-intercept (FI) propagation models. This is achieved through the use of bio-inspired optimization techniques, genetic algorithms (GA) and Particle Swarm Optimization (PSO). The study is contextualized in outdoor scenario, with particular emphasis on the Rio Negro Bridge, in Manaus, state of Amazonas, Brazil, with a measurement campaign focused on signal strength data. The goal is to characterize the average behavior of path loss in this specific environment. The results of this research include the identification of the most appropriate model coefficients (CI and FI) that fit the measured data, with the fitted CI and FI models having the smallest values of Root Mean Square Error (RMSE) of 4.28 dB and 4.14 dB respectively compared to classic path loss models from the literature, such as Okumura-Hata (7.55 dB), ECC-33 (8.54 dB) and Ericsson (16.9 dB). Consequently, the model adapted in this study has the potential to make a significant contribution to the strategic planning and implementation of 5G network expansion in the Amazon region.
The Internet of Things (IoT) demands an evolution of current networking protocols to support the requirements of its various application scenarios. One such example is the amendment IEEE 802.11ah (a.k.a. Wi-Fi HaLow) designed to operate in sub-GHz frequency bands and achieve higher signal coverage than typical Wi-Fi. Its Restricted Access Window (RAW) mechanism decreases channel contention by dividing stations into RAW groups that can be further divided into RAW slots. However, few works have addressed the impact of Rayleigh channel fading on IEEE 802.11ah networks. To study this problem, we extend a discrete-time Markov chain model to consider the impact of Rayleigh channel fading on the operation of IEEE 802.11ah networks. We validate the model via ns-3 simulations and propose grouping strategies to measure efficiency in terms of Jain’s fairness index. Additionally, we propose an expression to adjust the RAW slot duration based on the distance to the access point (AP).
We investigate a cascaded channel estimation approach for reconfigurable intelligent surface (RIS)-aided millimeter-wave (mmWave) multi-user multiple-input single-output communication systems. To mitigate the high cost and power consumption of analog-to-digital converters (ADCs) in mmWave base stations with many antennas and wide signal bandwidths, we propose a task-based quantization approach using identical low-resolution scalar ADCs that minimizes the channel estimation error for hybrid analog and digital architectures. Numerical results verify that the performance of the proposed channel estimation design can effectively approach that of a system operating with unlimited resolution ADCs, and outperform a system operating solely in the digital domain under identical bit-resolution constraints.
Super-directive antenna arrays (SDAAs) suffer from low realized gain caused by high impedance mismatches and low radiation efficiencies. Additionally, since the maximum radiation is directed only towards the end-fire, the practical applications of SDAAs have been historically restricted due to their limited coverage area. When classical phase shifters are employed for beam-steering, there is a significant decrease in the gain attributed to the amplified sensitivity of the realized gain to antenna feed currents. Therefore, innovative beamforming network (BFN) configurations are required to direct the beams while maintaining high realized gain in SDAAs. To address this fundamental challenge, we introduce a novel configuration employing parasitically loaded elements for beam-steering applications within a four-element array, aiming to achieve high realized gain. Firstly, the effect of the parasitic elements' positions on the array's beam-steering performance is examined. Subsequently, a reconfigurable BFN is introduced, incorporating adjustable passive circuits within the parasitic presented comparatively. The common belief that constrains the practical applications of SDAAs - that "maximum radiation is directed towards end-fire" - is challenged as a result. Testament to this, we achieve a 34% increase in gains compared to a uniform uncoupled array for the final loaded configuration, augmented with the ability to steer the beam towards any desired direction.
Mobile-edge computing (MEC) has been introduced as a promising paradigm to provide computing resources to resource-limited devices. Currently, most of the research consider static computation offloading and resource allocation in MEC, while ignoring the impact of mobility of mobile users (MUs). In this paper, we propose an algorithm to optimize the task offloading and resource allocation in Digital Twin (DT)-empowered MEC system in two steps. First, to obtain stable offloading choices, we propose an innovative algorithm for predicting stable association of MU and edge node (EN) pairs assisted by DT to enhance users’ experience. Second, taking into account the contention for resources and the constraints of offloading tasks, we construct an incentive mechanism based on reverse-auction and design an unit-revenue-based algorithm to maximize the overall revenue of MUs’ side. Simulation results demonstrate that the proposed algorithm can improve the task offloading success rate and the overall system utility, thus show the the effectiveness of our algorithm.
This article evaluates the performance of two artificial neural network (ANN) techniques, multilayer perceptron (MLP) and long short-term memory (LSTM), in predicting the level of mobile radio signals inside a tunnel environment, with data collected at the 5.8 GHz frequency, mainly relevant for communication with and between vehicles. This study investigates the effectiveness of ANN-based models compared to a traditional reference model, log-distance (LD), for predicting received signal level. By analyzing the results, it was possible to identify the relevant factors of the MLP and LSTM networks that must be considered when building such models.
We propose a probabilistic method to estimate the optimal phase shifts of a RIS in the absence of channel estimation. We formulate the problem with a bayesian approach and propose several algorithms to optimize the achievable rate. Numerical results compare the performances of the algorithm with well-known benchmarks. Our method outperforms those benchmarks on different codebooks both at high and low SNR.
The study of advanced architectures for reconfigurable intelligent surfaces (RIS) integrated sensing capabilities has been the focus of recent research in the past few years. A promising architecture is the so-called hybrid RIS (HRIS), which comprises hybrids of sensing and reflecting meta-atoms. This enables HRIS to have signal processing capabilities, which can be used to partially solve the channel estimation task. This work proposes a semi-blind approach to the HRIS-assisted MIMO wireless communication system by empowering the HRIS with joint symbol detection and channel estimation functions based on alternating least squares estimation methods. Simulation results indicate the remarkable performance of the proposed receivers in terms of normalized mean square error and symbol error rate, compared to solutions found in the literature, showing the efficacy of HRIS in MIMO systems and the potential of semi-blind methods in this context.
The performance of massive MIMO wireless communication links depends on the spatial statistics of the underlying wireless channel. While in literature, it is often assumed that MIMO channels are either independent or follow a correlation model according to the Kronecker structure, theory as well as measurements suggest that a more detailed view on the spatial dependency structure is required to design efficient MIMO communication links. The resource efficient design of massive MIMO systems is very relevant for the B5G and 6G generations of cellular and WiFi networks. In this paper, we propose a dependency model for the spatial channel gains in massive MIMO links. In particular, we show the limitations of the state of the art Kronecker-based correlation model, and propose a copula-based model. For two current beamforming strategies, selection combining and maximum ratio combining, we completely characterize the impact of the spatial dependency on the outage probability. In selection combining, the impact is clear, and less dependency improves outage performance. In maximum ratio combining, there is a dichotomy: for low outage probabilities (below 0.5), less dependency improves performance, while for high outage probabilities (above 0.5), more dependency improves performance. We compare our proposed model with channel measurements obtained from our massive MIMO prototyping system and show that our spatial dependency model can well describe the measurements. Finally, the predicted performance behavior is verified by the measurements. The results indicate that a fine granular model for channel dependency can better predict outage performance.
Unmanned aerial vehicle (UAV) communication is seen as an emerging technology in the future network, benefiting from its mobility and high quality channels such that the transmission efficiency of semantic information can be highly improved. Thus for task success probability improvement with fairness consideration, we provide a semantic throughput maximization design through jointly optimizing UAV trajectory and user scheduling, under mobility and scheduling constraints. To solve the complicated problem with immeasurable variables, we adopted the successive-hover-and-fly (SHF) structure such that the original problem is reformulated into one with limited number of variables. While the reformulated problem is still highly non-convex, we rigorous prove the convexity of the transmission rate in PLoS channels which helps in conducting a convex approximation and enables a high-performance iteration algorithm to obtain a high-quality solution. Numerical results demonstrate the high performance of proposed algorithm.
In this paper, two new correntropy-based data-selective algorithms using the prediction error method and proportionate principle are proposed for acoustic feedback cancellation (AFC). The proposed data selective approach is applied for the improved practical variable step size proportionate normalized least mean square algorithm (IPNLMS-IPVSS) and its novel tanh-based version. It is shown that the proposed data-selective algorithms can achieve close performance to the original algorithms at a reduced average numerical complexity.
