Machine learning (ML) algorithms have been widely used to improve the performance of telecommunications systems, including reconfigurable intelligent surface (RIS)-assisted wireless communication systems. The RIS can be considered a key part of the backbone of sixth-generation (6G) communication mainly due to its electromagnetic properties for controlling the propagation of the signals in the wireless channel. The ML-optimized (RIS)-assisted wireless communication systems can be an effective alternative to mitigate the degradation suffered by the signal in the wireless channel, providing significant advantages in the system’s performance. However, the variety of approaches, system configurations, and channel conditions make it difficult to determine the best technique or group of techniques for effectively implementing an optimal solution. This paper presents a comprehensive review of the reported frameworks in the literature that apply ML and RISs to improve the overall performance of the wireless communication system. This paper compares the ML strategies that can be used to address the RIS-assisted system design. The systems are classified according to the ML method, the databases used, the implementation complexity, and the reported performance gains. Finally, we shed light on the challenges and opportunities in designing and implementing future RIS-assisted wireless communication systems based on ML strategies.

Worldwide, heart disease is the leading cause of mortality. Cardiac auscultation, when conducted by a trained professional, is a non-invasive, cost-effective, and readily available method for the initial assessment of cardiac health. Automated heart sound analysis offers a promising and accessible approach to supporting cardiac diagnosis. This work introduces a novel method for classifying heart sounds as normal or abnormal by leveraging time-frequency representations. Our approach combines three distinct time-frequency representations—short-time Fourier transform (STFT), mel-scale spectrogram, and wavelet synchrosqueezed transform (WSST)—to create images that enhance classification performance. These images are used to train five convolutional neural networks (CNNs): AlexNet, VGG-16, ResNet50, a CNN specialized in STFT images, and our proposed CNN model. The method was trained and tested using three public heart sound datasets: PhysioNet/CinC Challenge 2016, CirCor DigiScope Phonocardiogram Dataset 2022, and another open database. While individual representations achieve maximum accuracy of ≈85.9%, combining STFT, mel, and WSST boosts accuracy to ≈99%. By integrating complementary time-frequency features, our approach demonstrates robust heart sound analysis, achieving consistent classification performance across diverse CNN architectures, thus ensuring reliability and generalizability.

Multiple input-multiple output (MIMO) is a key enabling technology for the next generation of wireless communication systems. However, one of the main challenges in the implementation of MIMO system is the complexity of the detectors when the number of antennas increases. This aspect will be crucial in the implementation of future massive MIMO systems. A flexible design can offer a convenient tradeoff between detection complexity and bit error rate (BER). Deep learning (DL) has emerged as an efficient method for solving optimization problems in different areas. In MIMO communication systems, neural networks can provide efficient and innovative solutions. This paper presents an efficient DL-based signal detection strategy for MIMO communication systems. More specifically, a preprocessing stage is added to label the input signals. The labeling scheme provides more information about the transmitted symbols for better training. Based on this strategy, two novel schemes are proposed and evaluated considering BER performance and detection complexity. The performance of the proposed schemes is compared with the conventional one-hot (OH) scheme and the optimal maximum likelihood (ML) criterion. The results show that the proposed OH per antenna (OHA) and direct symbol encoding (DSE) schemes reach a classification performance F1-score of 0.97. Both schemes present a lower complexity compared with the conventional OH and the ML schemes, used as references. On the other hand, the OHA and DSE schemes have losses of less than 1 dB and 2 dB in BER performance, respectively, compared to the OH scheme. The proposed strategy can be applied to adaptive systems where computational resources are limited.

Cardiovascular diseases (CVDs) remain the leading cause of morbidity worldwide. The heart sound signal or phonocardiogram (PCG) is the most simple, low-cost, and effective tool to assist physicians in diagnosing CVDs. Advances in signal processing and machine learning have motivated the design of computer-aided systems for heart illness detection based only on the PCG. The objective of this work is to compare the effects of using spectral and sparse features for a classification scheme to detect the presence/absence of a pathological state in a heart sound signal, more specifically, sparse representations using Matching Pursuit with multiscale Gabor time-frequency dictionaries, linear prediction coding, and Mel-frequency cepstral coefficients. This work compares the performance of PCGs classification applying features as a result of averaging the samples or the features for each PCG sound event when feeding a random forest (RF) classifier. For data balancing, random under-sampling and synthetic minority oversampling (SMOTE) methods were applied. Furthermore, we compare the Correlation Feature Selection (CFS) and Information Gain (IG) for the dimensionality reduction. The findings show a SE=93.17 %, SP=84.32 % and ACC=85.9 % when joining MP+LPC+MFCC features set with an AUC=0.969 showing that these features are promising to be used in heart sounds anomaly detection schemes.

This paper presents a comprehensive probabilistic analysis of packet loss in long-range (LoRa) networks, a vital aspect of low-power, wide-area communication systems increasingly employed in IoT applications. The proposed model integrates multiple critical factors, including packet arrival rates, transmission power levels, and the distance between transmitting nodes and the gateway. By incorporating these variables into a unified probabilistic framework, the model not only predicts packet loss and interference patterns but also provides insights into optimizing network parameters. Specifically, it focuses on determining the optimal transmission power required to balance energy efficiency and communication reliability. A distinctive feature of the analysis is its ability to adapt dynamically to varying network conditions, ensuring sustained performance even in environments with high node density or fluctuating traffic loads. The study also explores the interplay between transmission power and interference, demonstrating how careful calibration of power settings can significantly reduce packet collisions while conserving energy resources. The proposed framework not only advances theoretical understanding, but also offers actionable guidelines for network designers seeking to achieve high performance in resource-constrained environments.