International Journal of Scientific Research and Engineering Development

Nowadays, almost all the household, corporate, and academic networks are based on wireless infrastructure, and this dependency creates a great number of targets for potential attacks. There are many unauthorized access attempts, MAC address spoofing, packet sniffing, brute force attacks, and denial-of-service floods. The traditional, rule-based security measures of these networks cannot withstand them anymore. Besides, there is another important problem associated with wireless communication - inefficiency of the bandwidth usage. However, this issue becomes obvious only when users start complaining about it. Instead of tackling security and efficiency separately, this work suggests a solution that will tackle both of these problems simultaneously. Captured live traffic is analyzed with Scapy and cross-checked through protocol analysis in Wireshark, while Nmap scans network to discover hosts, identify OS type, and map open ports/services. The behavioral features extracted through the process become the input for five classifiers: Random Forest, Decision Tree, and Support Vector Machine classifiers, as well as XGBoost and Isolation Forest models, which can detect anomalies that are not present in the labeled training data. The deviation from the device-specific baseline behavior can be detected due to the device-specific monitoring that was implemented, and, as a result, there will be the risk assessment for each individual device in the form of a risk score. This risk score serves as the basis for the recommendation engine (e.g. block device, require WPA3, disable WPS, change credentials), and, at the same time, as the input to the adaptive optimization layer for bandwidth allocation, channel selection, and traffic prioritization. Live Flask web application will be the user interface for this system. The majority of the research papers deal with the issues of security or performance, and they rarely solve both problems at once - this paper states that the feedback loop between security and performance should exist. It provides the architecture and a sixteen-week roadmap to develop it, and discusses how such a system can be assessed in the enterprise, healthcare, IoT, and smart home applications.

Adithya K, Darshan Chandrakanth Kulkarni, Ekanth R, Harsha A, Dr. R. Sivakumar, "AI-Based Intelligent Wi-Fi Security and Adaptive Optimization" International Journal of Scientific Research and Engineering Development, V9(4): Page(1165-1169) July-August 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.