Recent Papers

The rapid expansion of Internet of Things (IoT) infrastructure has created an interconnected ecosystem of billions of intelligent devices, facilitating automated operations across healthcare, manufacturing, agriculture, transportation, and urban planning sectors. Despite the substantial benefits IoT provides, the extensive attack vectors, limited computational capabilities, and diverse nature of IoT systems create significant security challenges that make these devices attractive targets for cybercriminals. Threat actors leverage these weaknesses to execute various attacks including Distributed Denial of Service (DDoS) operations, information theft, malicious code insertion, identity spoofing, and unauthorized system access, potentially causing critical failures in mission-critical applications.This research addresses these security concerns by developing and deploying an advanced cyber-threat detection framework specifically optimized for IoT ecosystems. The methodology encompasses gathering network and device telemetry from IoT endpoints and examining this data through machine learning algorithms to detect irregular patterns that suggest potential security threats. The system framework incorporates components for data cleaning, feature identification, algorithm training, and continuous threat monitoring.Various supervised learning methodologies including Random Forest, Support Vector Machine (SVM), and Decision Trees, along with unsupervised techniques like K-Means Clustering and Autoencoders, are assessed for their capability in differentiating between legitimate and hostile activities. System validation employs industry-standard datasets including UNSW-NB15, CICIDS2017, and Bot-IoT, with performance evaluation using metrics such as accuracy, precision, recall, F1-score, and false positive rates. Experimental findings show that the developed framework can identify diverse attack patterns with superior accuracy and minimal resource consumption, making it appropriate for implementation in resource-limited IoT deployments.Given the increasing frequency of sophisticated attacks, there is an urgent requirement for intelligent, responsive, and analytics-driven solutions capable of categorizing and forecasting DDoS attacks in real-time. Machine learning, as a branch of artificial intelligence, has proven to be an effective methodology for addressing this challenge. This research introduces an innovative machine learning-based classification and prediction framework for DDoS attack detection.

Copyright © 2025 B. Vijay kumar. This is an open access article distributed under the Creative Commons Attribution License.