Deep learning (DL) is a new machine learning (ML) methodology that has found successful implementations in many application domains. However, its usage in communications systems has not been well explored. This paper investigates the use of the DL in modulation classification, which is a major task in many communications systems. The DL relies on a massive amount of data and, for research and applications, this can be easily available in communications systems. Furthermore, unlike the ML, the DL has the advantage of not requiring manual feature selections, which significantly reduces the task complexity in modulation classification. In this paper, we use two convolutional neural network (CNN)-based DL models, AlexNet and GoogLeNet. Specifically, we develop several methods to represent modulated signals in data formats with gridlike topologies for the CNN. The impacts of representation on classification performance are also analyzed. In addition, comparisons with traditional cumulant and ML-based algorithms are presented. Experimental results demonstrate the significant performance advantage and application feasibility of the DL-based approach for modulation classification.

About 26 million people worldwide experience its effects each year. Both cardiologists and surgeons have a tough time determining when heart failure will occur. Classification and prediction models applied to medical data allow for enhanced insight. Improved heart failure projection is a major goal of the research team using the heart disease dataset. The probability of heart failure is predicted using data mined from a medical database and processed by machine learning methods. It has been shown, through the use of this study and a comparative analysis, that heart disease may be predicted with high precision. In this study, researchers developed a machine learning model to improve the accuracy with which diseases like heart failure (HF) may be predicted. To rank the accuracy of linear models, we find that logistic regression (82.76 percent), SVM (67.24 percent), KNN (60.34 percent), GNB (79.31 percent), and MNB (72.41) perform best. These models are all examples of ensemble learning, with the most accurate being ET (70.31%), RF (87.03%), and GBC (86.21%). DT (ensemble learning models) achieves the highest degree of precision. CatBoost outperforms LGBM, HGBC, and XGB, all of which achieve 84.48% accuracy or better, while XGB achieves 84.48% accuracy using a gradient-based gradient method (GBG). LGBM has the highest accuracy rate (86.21 percent) (hypertuned ensemble learning models). A statistical analysis of all available algorithms found that CatBoost, random forests, and gradient boosting provided the most reliable results for predicting future heart attacks.

Memory Denial of Service (M-DoS) attacks refer to a class of cyber-attacks that aim to exhaust the memory resources of a system, rendering it unavailable to legitimate users. This type of attack is particularly dangerous in cloud computing environments, where multiple users share the same resources. Detection and mitigation of M-DoS attacks in real-time is a challenging task, as they often involve a large number of low-rate requests, making it difficult to distinguish them from legitimate traffic. Several real-time detection schemes have been proposed to identify and mitigate M-DoS attacks in cloud computing environments. These schemes can be broadly classified into two categories: signature-based and anomaly-based detection. Signature-based detection methods rely on the identification of specific patterns or characteristics of known M-DoS attack techniques, while anomaly-based detection methods identify abnormal behaviour that deviates from the normal pattern of usage. This study presents a hybrid model for real-time detection of cloud and MDOS attacks using SVM-KNN-LR. The dataset used in this study was collected from various sources and pre-processed to extract relevant features for attack detection. A feature selection process was also applied to identify the most important features for attack detection. The hybrid model achieved an accuracy of 96%, outperforming other individual models such as SVM, KNN, LR, Naive Bayes, Decision Trees, Extra Trees, Bagging Trees, and Random Forests. Confusion matrices were also used to evaluate the performance of each model. In the discussion section, we examined the performance of the hybrid model in detecting MDOS attacks and found that it had a high precision score of 0.97. However, the recall score was lower at 0.87, indicating that the model was not able to detect all instances of MDOS attacks.

The rapid advancement of terahertz (THz) communication systems has positioned this technology as a key enabler for next-generation telecommunication networks, including 6G, secure communications, and hybrid wireless-optical systems. This review comprehensively analyzes THz communication, emphasizing its integration with free-space optical (FSO) systems to overcome conventional bandwidth limitations. While THz-FSO technology promises ultra-high data rates, it is significantly affected by atmospheric absorption, particularly absorption beyond 500 GHz, where the attenuation exceeds 100 dB/km, which severely limits its transmission range. However, the presence of a lower-loss transmission window at 680 GHz provides an opportunity for optimized THz-FSO communication. This paper explores recent developments in high-power THz sources, such as quantum cascade lasers, photonic mixers, and free-electron lasers, which facilitate the attainment of ultra-high data rates. Additionally, adaptive optics, machine learning-based beam alignment, and low-loss materials are examined as potential solutions to mitigating signal degradation due to atmospheric absorption. The integration of THz-FSO systems with optical and radio frequency (RF) technologies is assessed within the framework of software-defined networking (SDN) and multi-band adaptive communication, enhancing their reliability and range. Furthermore, this review discusses emerging applications such as self-driving systems in 6G networks, ultra-low latency communication, holographic telepresence, and inter-satellite links. Future research directions include the use of artificial intelligence for network optimization, creating energy-efficient system designs, and quantum encryption to obtain secure THz communications. Despite the severe constraints imposed by atmospheric attenuation, the technology’s power efficiency, and the materials that are used, THz-FSO technology is promising for the field of ultra-fast and secure next-generation networks. Addressing these limitations through hybrid optical-THz architectures, AI-driven adaptation, and advanced waveguides will be critical for the full realization of THz-FSO communication in modern telecommunication infrastructures.

As criminal activity increasingly relies on digital devices, the field of digital forensics plays a vital role in identifying and investigating criminals. In this paper, we addressed the problem of anomaly detection in digital forensics data. Our objective was to propose an effective approach for identifying suspicious patterns and activities that could indicate criminal behavior. To achieve this, we introduce a novel method called the Novel Support Vector Neural Network (NSVNN). We evaluated the performance of the NSVNN by conducting experiments on a real-world dataset of digital forensics data. The dataset consisted of various features related to network activity, system logs, and file metadata. Through our experiments, we compared the NSVNN with several existing anomaly detection algorithms, including Support Vector Machines (SVM) and neural networks. We measured and analyzed the performance of each algorithm in terms of the accuracy, precision, recall, and F1-score. Furthermore, we provide insights into the specific features that contribute significantly to the detection of anomalies. Our results demonstrated that the NSVNN method outperformed the existing algorithms in terms of anomaly detection accuracy. We also highlight the interpretability of the NSVNN model by analyzing the feature importance and providing insights into the decision-making process. Overall, our research contributes to the field of digital forensics by proposing a novel approach, the NSVNN, for anomaly detection. We emphasize the importance of both performance evaluation and model interpretability in this context, providing practical insights for identifying criminal behavior in digital forensics investigations.