Real-Time Anomaly Detection System (RADS): A Lightweight, Database-Free Approach for System Performance and Security Monitoring

Modern computing systems require continuous monitoring to maintain stability, performance efficiency, and security. Sudden anomalies such as abnormal CPU utilization, excessive memory consumption, or unusual network activity may indicate performance degradation or malicious behavior. Traditional anomaly detection systems often depend on centralized databases, historical log analysis, or computationally intensive machine learning models, which introduce processing delays and increase system overhead. This paper proposes a Real-Time Anomaly Detection System (RADS), a lightweight and database-free framework designed for instant detection of abnormal system behavior. The system continuously monitors CPU, RAM, and network usage using Python-based monitoring techniques and applies threshold-based and adaptive detection mechanisms. Upon anomaly detection, RADS immediately generates alerts through Discord, enabling rapid response without manual supervision. Experimental evaluation demonstrates that the proposed system achieves accurate real-time monitoring with minimal resource consumption, making it suitable for local systems, academic environments, and small-scale servers. The proposed system is optimized for Ubuntu-based environments but can be extended to other operating systems with minimal changes. Experimental evaluation demonstrates that RADS delivers accurate, real-time anomaly detection with minimal resource overhead. By combining performance monitoring and security awareness in a single, accessible framework, RADS contributes a practical solution for real-time system reliability and protection