Detecting attacks on a mesh network using anomaly detecting techniques

Presentation Type

Abstract

Faculty Advisor

Christopher Leberknight

Access Type

Event

Start Date

25-4-2025 1:30 PM

End Date

25-4-2025 2:29 PM

Description

Mesh networks are widely used for their robustness and flexibility, connecting multiple nodes in decentralized architectures. However, they are vulnerable to various security threats, including Denial of Service (DoS) and Distributed Denial of Service (DDoS) attacks, which can severely disrupt network operations. This research explores the application of anomaly detection techniques to detect such attacks within mesh networks. By continuously monitoring traffic flow, node behavior, and communication patterns, anomaly detection systems can identify deviations from normal network activity, which may indicate the presence of malicious attacks. The research examines various approaches to anomaly detection, such as statistical analysis, machine learning, and flow-based analysis, highlighting their effectiveness in detecting sudden traffic surges, abnormal node behavior, and irregular request frequencies—common indicators of DoS/DDoS attacks. We discuss how these techniques can proactively identify attacks in real-time, allowing for quicker response and mitigation strategies to ensure network resilience and minimize service disruption. The results demonstrate the potential of anomaly detection in enhancing the security and reliability of mesh networks against such attacks.

Comments

Poster presentation at the 2025 Student Research Symposium.

This document is currently not available here.

Share

COinS
 
Apr 25th, 1:30 PM Apr 25th, 2:29 PM

Detecting attacks on a mesh network using anomaly detecting techniques

Mesh networks are widely used for their robustness and flexibility, connecting multiple nodes in decentralized architectures. However, they are vulnerable to various security threats, including Denial of Service (DoS) and Distributed Denial of Service (DDoS) attacks, which can severely disrupt network operations. This research explores the application of anomaly detection techniques to detect such attacks within mesh networks. By continuously monitoring traffic flow, node behavior, and communication patterns, anomaly detection systems can identify deviations from normal network activity, which may indicate the presence of malicious attacks. The research examines various approaches to anomaly detection, such as statistical analysis, machine learning, and flow-based analysis, highlighting their effectiveness in detecting sudden traffic surges, abnormal node behavior, and irregular request frequencies—common indicators of DoS/DDoS attacks. We discuss how these techniques can proactively identify attacks in real-time, allowing for quicker response and mitigation strategies to ensure network resilience and minimize service disruption. The results demonstrate the potential of anomaly detection in enhancing the security and reliability of mesh networks against such attacks.