India: Intruder Node Detection and Isolation Action in Mobile Ad Hoc Networks Using Feature Optimization and...

Area of Science:

• Computer Science
• Network Security
• Mobile Computing

Background:

• Mobile ad hoc networks (MANETs) lack centralized security, making them vulnerable to malicious attacks.
• Attacks in MANETs can lead to significant data loss, energy depletion, and reduced throughput.
• Existing security mechanisms may not be efficient or lightweight enough for dynamic MANET environments.

Purpose of the Study:

• To propose a novel, lightweight mechanism for detecting and isolating intruder nodes in MANETs.
• To enhance the security and reliability of MANETs against malicious activities.
• To evaluate the effectiveness of the proposed mechanism in terms of security and performance metrics.

Main Methods:

• Feature extraction of direct and indirect trust metrics from each node.
• Feature optimization using the Particle Swarm Optimization (PSO) algorithm.
• Classification of optimized features using a Neural Networks (NN) classifier to identify intruder nodes.
• Implementation of an Intruder Node Detection and Isolation Action (INDIA) mechanism.

Main Results:

• The INDIA mechanism successfully identifies intruder nodes within the MANET.
• Performance evaluation considered packet delivery success rate, communication delay, and energy consumption.
• The proposed method demonstrates effectiveness in detecting and isolating malicious nodes.

Conclusions:

• The INDIA mechanism offers a lightweight and effective solution for enhancing MANET security.
• The combination of feature extraction, PSO optimization, and NN classification provides robust intruder detection.
• The study highlights the importance of trust-based mechanisms for securing decentralized networks.