Simple Network Management Protocol (SNMP) is extensively utilized for monitoring and managing computer networks. This paper synthesizes and reviews recent research on harnessing SNMP data for developing effective network security solutions driven by artificial intelligence and machine learning techniques. The comprehensive device-level visibility and timely data provided by SNMP offers significant potential to enable high-performance intrusion detection systems, security analytics, and real-time monitoring capabilities to protect against evolving cyber threats. However, inherent vulnerabilities in versions like SNMP v1/v2c persist, especially due to default community strings allowing uncontrolled access. A number of techniques like newer SNMPv3 standard have been proposed to improve security through encryption, authentication, and fine-grained access control. Additionally, general best practices around firewalling, traffic monitoring, and compliance audits are recommended for robust SNMP deployment. The literature reviewed demonstrates SNMP’s capabilities in enabling anomaly detection with over 99% accuracy by applying supervised learning methods like random forests and neural networks to SNMP Management Information Base (MIB) data. The network-centric visibility offered by SNMP's extensive statistics on traffic configurations and behaviors provides an invaluable advantage over packetlevel data for developing intelligent detection models. Beyond intrusion detection systems, SNMP data has also shown significant utility for security analytics like attack pattern discovery, bandwidth modeling, and large-scale threat intelligence by aggregating data from distributed network devices. Further research opportunities exist in areas like feature engineering, model optimization, and continued evaluation of evolving attacks patterns. In conclusion, with appropriate secure configurations, SNMP provides the comprehensive and real-time data needed by modern artificial intelligence security applications in a scalable manner due to its device-centric focus, standardized schema, and ubiquitous deployment

Portable Document Format (PDF) is one of the most widely used files types worldwide in data exchange, this has encourage hackers to utilize such files to spread any malicious content through PDF, utilizing different methods and techniques to accomplish that, on the other hand, security researches kept trying to improve detection methods to cope up to the rapidly increasing number of malwares daily, one of the commonly used detection technique nowadays is by utilizing artificial intelligence and Machine learning classificat; thision to help detecting PDF Malwares, in this paper, we utilize machine learning classifier Random Forest on a newly released PDF Malware dataset CIC-Evasive-PDFMal2022 to achieve the main goal of detecting malicious PDF documents, results showing a detection accuracy of around 99.5%

The digital society is an outcome of the Internet which has nearly made everything connected and accessible no matter where or when. Nevertheless, despite the fact that conventional IP networks are complicated and very hard to manage, they are still widely adopted. The already established policies make the network configuration/reconfiguration a complex process that reacts to errors, load, and modifications. The prevailing networks are vertically integrated which makes things more and more complicated: Data planes and control are strapped together. Software-defined networking is a model that is meant to solve this issue by splitting the vertical integration and detaching the network’s control logic from the implicit routers and switches; this could be achieved by reinforcing centralization of network control and making the network programmable. In this work, we worked to implement MPLS networks with SDN, to enhance the traffic engineering over the network, and to minimize the network delay and latency, with minimum cost using three of the different SDN networks. The experiment results showed the advantage of the proposed approach for reducing the network delay, comparing with previous studies. Where the average of network delay in our approach reaches to 3.01 milliseconds.

The growing shift toward Internet of Things (IoT)-based solutions expands the attack surface of systems by connecting an extensive network of heterogeneous devices and technologies. The heterogeneity of IoT devices and the scale of the network often make conventional security measures impractical. Therefore, recent research efforts have focused on machine learning (ML)-based device-agnostic IoT security solutions. However, most proposed solutions have focused solely on detecting malicious traffic in IoT networks. While this is important, further information about the attack is needed to provide a comprehensive defense before, during, and after the incident. To address this gap, we proposed an anomaly-based intrusion detection model that detects malicious traffic and identifies the attack category and subcategory with high accuracy, recall, precision, and F1-score. The proposed intrusion detection model is an ensemble model that integrates multiple ML models to produce more robust and reliable detections. For evaluation, we performed three sets of experiments: (1) a binary classification to detect malicious traffic; (2) a multi-class classification to detect the attack category; and (3) a multi-class classification to detect the attack subcategory. In the experiments, our model achieved an accuracy, recall, precision, and F1-score of 100% for the binary detection; a precision, recall, and F1-score of 99% and an accuracy of 100% for the multi-class category classification; and an accuracy, recall, precision, and F1-score of 88% for the multi-class subcategory classification. These results suggest that the proposed model can reliably detect anomalies in IoT devices.