In this paper, we derive for the first time the analytical expressions of the exact Cramér-Rao lower bounds (CRLBs) of the carrier frequency and the carrier phase from square quadrature amplitude modulated (QAM) signals, assuming the noise power and the signal amplitude to be completely unknown. The signal is assumed to be corrupted by additive white Gaussian noise (AWGN). The main contribution of this paper consists in deriving the analytical expressions for the non-data-aided (NDA) Fisher information matrix (FIM) for higher-order square QAM-modulated signals. We prove that the problem of estimating the synchronization parameters is separable from the one of estimating the signal and the noise powers by showing that the FIM is block diagonal. Besides, we show analytically that the phase CRLB is higher than the frequency CRLB, implying that it is much easier to estimate the frequency than the distortion phase. It will be seen that the CRLBs differ widely from one modulation order to another in the medium SNR range. The newly derived expressions corroborate previous attempts to numerically or empirically compute the considered CRLBs as well as their asymptotical expressions derived only in special SNR regions.

Stroke is a leading cause of death and disability worldwide, requiring accurate and early prediction to ensure timely medical intervention. This study proposes a hybrid system that combines optimal feature selection and advanced classification techniques to improve stroke prediction performance. We used a publicly available Harvard Stroke Prediction Data Warehouse dataset, applying multiple feature selection methods: ANOVA, chi‐square, mutual information classification, and analysis of variance to identify relevant features. Five classifiers were examined: Random Forest (RF), K‐Nearest Neighbors (KNN), Decision Tree (DT), XGBoost, and Multilayer Perceptron (MLP). MLP was also used for feature selection through its internal representation learning capabilities. The parameters were fine‐tuned using GridSearchCV. The most effective configuration used selected features from a RF with MLP as a classifier, achieving 99.86% accuracy, 1.00% recall, 99.73% precision, and an F1 score of 99.86%. Compared with existing state‐of‐the‐art models, our proposed system demonstrates superior performance. This approach enables earlier and more accurate stroke detection, supporting healthcare providers in delivering personalized and proactive care to at‐risk individuals.

The rise of cyberattacks has necessitated the development of effective malware detection mechanisms. Deep learning, with its ability to learn complex features from raw data, has been widely used for this purpose. This paper presents a two-fold contribution to the field of malware detection using deep learning. Firstly, we use seven pretrained CNN models to classify malware images from the Malimg dataset. While these models achieved high accuracy, they presents low precision and F1-score, indicating that they were prone to false positives and false negatives. Additionally, these pre-trained models are susceptible to overfitting, which is a common issue with transfer learning. To overcome this limitation, we propose a custom CNN model consisting of six layers, trained from scratch on the Malimg dataset. To address potential issues like over-fitting and data imbalance when training models from scratch, we used regularization techniques such as L1 or L2 regularization, dropout. Our proposed custom CNN model outperformed the pretrained models, achieving best/average accuracy values over 5 trials of 100%/98.26%. We also found that our custom model was less susceptible to over-fitting and adversarial attacks. Our proposed approach provides a promising solution to the problem of limitations in using pre-trained models for malware detection.

Breast cancer is a prevalent and serious disease that affects many women around the world every year. Detecting breast cancer early is crucial for improving survival rates and treating the illness effectively. Researchers are exploring various methods, including neural networks and machine learning, to assist in detecting the disease. However, due to limited data availability, leveraging pre-trained models trained on diverse image datasets has become a common practice. This article introduces a novel approach to identifying breast cancer that involves the utilization of a deep learning model utilizing the ResNet50 framework, coupled with heat mapping and gradient-weighted class activation mapping (Grad-CAM). The suggested method was primarily assessed using the FDDM dataset of Subtracted Contrast Enhanced Spectral Mammography (CDD-CESM) images. The outcomes from this model were then contrasted with those of five other well-known models: VGG16, VGG19, MobileNetV2, EfficientNet-B7, and standard ResNet50. The newly proposed model yielded an accuracy of 0.8920, which was better than the other models. Additionally, Grad-CAM showed nearly flawless feature extraction in a breast cancer classification assignment. In the discussion section, the suggested method was utilized with the MIAS dataset to ensure thoroughness and scalability, and to allow for comparison with prior research. The results demonstrated the effectiveness of the suggested approach, with an accuracy of 0.9830 achieved on the MIAS dataset, surpassing previous works. This study significantly enhances the improvement of breast cancer detection through the integration of deep learning, the ResNet50 architecture, and visualization methods including heat maps and Grad-CAM.

Energy saving strategies pose challenges for wireless Sensor Networks communication, as the nodes have limited energy resources and are battery powered. Thus, there is a strong interest in developing energy efficient transmission techniques. In this paper, we present a new approach to minimize the energy consumption of clustered wireless sensor networks, while achieving the required level of reliability. We consider clusters where sensors transmit monitored data to nearby cluster-head using cooperative communication. The goal of our approach is to minimize the transmit power energy based on Bit Error Probability (BEP) criterion for selective digital relaying (SDR) schemes whose closed form expression was recently derived in previous works. Therefore, the proposed scheme can efficiently allocate the least transmission power among the source and the relays that satisfy the predefined level of reliability while decreasing the energy consumption of the node. Simulation results show that our relay selection algorithm achieves significant energy savings (more than 50% compared with direct transmissions for moderate SNR values and up to 90% in high SNR regions).