Impact Analysis of Malicious IoT Node Placement for Blackhole Attack Detection and Mitigation using Deep Learning-Based Trust Evaluation

• Published in: Indian journal of science and technology Vol. 18; no. 32; pp. 2581 - 2593
• Main Authors: Shameer, M, Rutravigneshwaran, P
• Format: Journal Article
• Language: English
• Published: 23.08.2025
• ISSN: 0974-6846; 0974-5645
• DOI: 10.17485/IJST/v18i32.1109

Objectives: The objective of this study is to create a deep learning-based trust evaluation mechanism that can detect blackhole incidents and investigate how faulty node positions impact IoT network performance. Methods: Six LoWPAN networks competing with the RPL protocol were simulated on IPv6 with Contiki OS and the Cooja simulator. Contiki is a light-weight operating system suitable for low-power devices. The standard simulator Cooja is widely used for RPL implementations, which allows testing and simulating networks for IoT devices. We simulated three models incorporating both internal and external hostile nodes. Using a deep learning-based Multilayer Perceptron (MLP) trust evaluation system, black hole assaults were detected through network metrics implementation, packet delivery ratio, latency, throughput, and power utilization. Findings: Our results emphasize the critical importance of implementing effective security mechanisms into use to locate and reduce internal black hole attacks. IoT systems need these mechanisms to maintain energy efficiency, network reliability, and overall integrity, especially in cases where reliability is crucial. Our research findings support the enhancement of security in RPL-based IoT systems and help researchers mitigate threats in an effective way. Novelty: This study presents significant outcomes, including Packet Delivery Ratio (PDR) of 94.44% and throughput of 87,406.38 bps, comparable to scenario S-II, which achieved a PDR of 69.23% and a throughput of 2,463.33 bps using a deep learning-based trust system concurrently with IoT network performance, particularly in the presence of core and outward blackhole occurrences. Keywords: Blackhole attacks, Deep learning, IoT security, RPL, Trust framework