Effectively detecting intrusions in the computer networks still remains problematic. This is because cyber attackers are changing packet contents to disguise the intrusion detection system (IDS) recently. Besides, everyday a lot of new devices are added to the computer networks. These new devices are also raising security issues in the computer networks. To effectively manage the computer network flows and provide the security in advance; the components of the IDSs, the approaches and technologies that are used, the nature of the attacks, and the tools that are used needs to be examined deeply. This paper discusses intrusion detection technologies, methodologies, and approaches and also investigates new attack types, protection mechanisms, and recent scientific studies that have been made in this area. In addition, available datasets, well-known IDS tools, and advantages and disadvantages of particular IDSs are explained deeply. We believe that this scientific review study presents a road map for researchers and industry employees who focus on IDSs.

These days, cloud computing is one of the most promising technologies to store information and provide services online efficiently. Using this rapidly developing technology to protect computer-based systems from cyber-related attacks can bring many advantages over traditional protection schemes. The protected assets can be any computer-based systems such as cyber-physical systems (CPS), critical systems, desktop and laptop computers, mobile devices, and Internet of Things (IoT). Malicious software (malware) is any software which targets the computer-based system to launch cyber-attacks to threaten the integrity, confidentiality and availability of the data. To detect the massively growing malware attacks surface, we propose an intelligent behavior-based detection system in the cloud environment. The proposed system first creates a malware dataset on different virtual machines which identify distinctive features efficiently. Then, selected features are given to the learning-based and rule-based detection agents to separate malware from benign samples. Totally, 10,000 program samples have been analyzed to evaluate the performance of the proposed system. The proposed system can detect both known and unknown malware efficiently with high detection and accuracy rate. Besides, the proposed method results have outperformed the leading methods’ results in the literature. Our evaluation results show that the proposed algorithms along with machine learning (ML) classifiers achieve 99.8% detection rate, 0.4% false positive rate, and 99.7% accuracy. Our proposed system and algorithms may assist those who would like to develop a novel malware detection system in the cloud environment.

Detecting anomalies in human activities is increasingly crucial today, particularly in nuclear family settings, where there may not be constant monitoring of individuals’ health, especially the elderly, during critical periods. Early anomaly detection can prevent from attack scenarios and life-threatening situations. This task becomes notably more complex when multiple ambient sensors are deployed in homes with multiple residents, as opposed to single-resident environments. Additionally, the availability of datasets containing anomalies representing the full spectrum of abnormalities is limited. In our experimental study, we employed eight widely used machine learning and two deep learning classifiers to identify anomalies in human activities. We meticulously generated anomalies, considering all conceivable scenarios. Our findings reveal that the Gated Recurrent Unit (GRU) excels in accurately classifying normal and anomalous activities, while the naïve Bayes classifier demonstrates relatively poor performance among the ten classifiers considered. We conducted various experiments to assess the impact of different training–test splitting ratios, along with a five-fold cross-validation technique, on the performance. Notably, the GRU model consistently outperformed all other classifiers under both conditions. Furthermore, we offer insights into the computational costs associated with these classifiers, encompassing training and prediction phases. Extensive ablation experiments conducted in this study underscore that all these classifiers can effectively be deployed for anomaly detection in two-resident homes.

The Internet of Robotic Things (IoRT) is revolutionizing industries by enabling autonomous, AI-driven robotic systems to perform complex and collaborative tasks, such as precision agriculture, disaster response, and logistic shipping operations. However, integrating AI into IoRT introduces significant challenges, including security vulnerabilities, adversarial attacks, data integrity risks, and operational disruptions in dynamic and high-stakes environments. This paper addresses these challenges by integrating and enhancing the NIST AI Risk Management Framework (AI RMF) for IoRT systems, providing a structured approach to identify, assess, and mitigate risks specific to IoRT ecosystems. We introduce a novel Large Language Model (LLM)-based approach for translating natural language commands into secure and precise robotic operations, enabling seamless collaboration and enhancing safety and reliability in mission-critical scenarios. Using a flood recovery scenario in precision agriculture, we demonstrate the practical application of these solutions, where swarm robots equipped with AI inference engines collaborate to navigate hazards, locate individuals, assess infrastructure damage, and mitigate risks. A comprehensive threat analysis is presented, mapping identified vulnerabilities to the NIST AIRMF, and tailored security controls are proposed to mitigate these threats effectively. We propose critical enhancements to the framework, including advanced quantitative risk assessment methods, subsystem governance strategies for interconnected IoRT networks, and robust auditing mechanisms to address unique IoRT-specific challenges. This work establishes a robust foundation for aligning AI governance frameworks with the complex and dynamic demands of IoRT systems. By addressing security, operational, and ethical considerations, it fosters secure, efficient, and trustworthy deployment across diverse applications, paving the way for sustainable and impactful IoRT innovations.