A System for Monitoring and Analysing LAN Network Traffic

Abstract

A network utilizing a LAN Telemetry System improves the performance and security of the network by offering comprehensive and continuous visibility and intelligence into the operation of the LAN to ensure a healthy and secure environment. This research explores the design, implementation, and evaluation of a scalable Local Area Network (LAN) telemetry system aimed at enhancing network performance and security. It addresses the complexities of data transfer in modern business environments, emphasizing the need for real-time monitoring and anomaly detection. The proposed work utilizes a three-tier architecture for data acquisition, processing, and visualization, incorporating tools such as libpcap for packet capturing and InfluxDB for time-series data storage. Evaluations indicate that the system achieves a high true positive rate of 98.5% in detecting security threats, while maintaining a low false positive rate below 2.5%. The study highlights the effectiveness of statistical anomaly detection in identifying potential security breaches and optimizing network performance, albeit acknowledging challenges in adapting to wider networks and encrypted communications. Future improvements may include machine learning applications to enhance threat detection capabilities.