Using Ensemble Learning Methodology for Predicting Reconnaissance in Campus Area Network

In the Campus Area Network (CAN) security is an important concern in the designed network. The reconnaissance domain of a cyberattack, allows attackers learn more about the system or network they are trying to attack. This first step of exploring, which is also called “cyber reconnaissance” or “cyber scouting,” lets threat attacker find weak spots, make a map of the network's design, and gather data they will need to launch more complex attacks. Early detection helps security teams stop harmful activity before it becomes more sophisticated and dangerous. It offers an important chance to strengthen defenses, fix vulnerabilities, and shore up vulnerabilities. This preventive strategy decreases cyber threats, illegal access, data breaches, and other cyber problems. Advanced threat intelligence, monitoring technologies, and anomaly detection help firms notice and react to cyber-attacks faster. To address these issues, a Simargl2022 dataset experiment was conducted. This dataset identified several dangers, prompting a port scanning attack study. This research employs an ensemble machine learning approach to detect port scanning attacks, hence enhancing network security and performance. To enhance the algorithm, ROC-AUC, accuracy, recall, and F1-Score are rigorously examined. The study discovered a high ability to detect attacks early on, preventing malicious groups from getting access. This proactive security keeps attackers out of the infrastructure and blocks dangerous data. The algorithm's network security improvements are shown by ROC-AUC, accuracy, recall, and F1-Score. This research demonstrates the algorithm's capabilities and emphasizes its significance in averting emerging cybersecurity threats.