International Journal of Mathematics And Computer Research

Abstract

The rapid growth of computing, a decentralized paradigm extending cloud computing capabilities closer to end users, has introduced new vulnerabilities and security challenges on file intrusion. Traditional Intrusion Detection Systems (IDS) often fall short in addressing the unique architecture. This study addresses the problem by developing a secured based intrusion detection model tailored for fog computing using the Naïve Bayes Algorithm. The methodology used in this work is object-oriented analysis and design methodology (OOADM) and python programming language is used to implement the model. The system integrates TensorFlow and OpenCV resources for robust data processing and real-time analysis. The model utilizes a policy-driven approach to handle security questions, ensuring compliance with predefined rules and standards. The results demonstrated the model’s capabilities to detect various types of intrusions and protection of data, such as unauthorized access attempts, with minimal computational overhead. The integration of Naïve Bayes classifier in a Fog environment is proven to be efficient and reliable, offering a scalable solution for real-time threat detection. The model was trained and tested using an 80/20 data split, achieving high accuracy and low latency in file intrusion detection. The model successfully identified and classified potential threats within Fog network, achieving an impressive accuracy of 97%. Its effectiveness was further evaluated through F1-score, which stood at 0.92, indicating a robust balance between precision and recall. This real-time detection is able to protect data from unauthorized parties from having access to files on the cloud.