Design and Development of a Cooperative Intrusion Detection System for Mobile Ad Hoc Networks

Abstract

Mobile Ad Hoc Networks (MANETs) are decentralized, self-configuring networks consisting of mobile nodes connected wirelessly without fixed infrastructure. Due to their dynamic topology, open medium, and lack of centralized control, MANETs are vulnerable to various security threats and attacks, making intrusion detection essential for maintaining network integrity. This paper presents the design and development of a Cooperative Intrusion Detection System (CIDS) tailored specifically for MANETs to enhance detection accuracy and network security. The proposed CIDS combines local intrusion detection modules on each node with a cooperative mechanism that enables nodes to share alerts and audit information. This cooperation leverages distributed monitoring to identify malicious behavior such as blackhole, wormhole, and denial-of-service (DoS) attacks effectively, overcoming limitations of isolated detection approaches. The study details the system architecture, incorporating anomaly-based and signature-based detection techniques to balance detection rates and false positives. The cooperative framework uses a trust management scheme to evaluate node reliability, preventing compromised nodes from spreading false alarms. Simulation experiments were conducted using NS-2 with various attack scenarios and node mobility patterns. Performance metrics, including detection rate, false alarm rate, network overhead, and energy consumption, were analyzed. Results show that the cooperative approach significantly improves detection accuracy compared to standalone IDS solutions, with acceptable communication overhead and energy use. This research contributes a scalable, lightweight intrusion detection framework suitable for resource-constrained MANET environments. The system’s modular design allows easy adaptation to evolving threats and network conditions. Future work will focus on integrating machine learning techniques for enhanced anomaly detection and expanding the trust model for dynamic environments.