Trust-Based Solutions: A Novel Approach to Lightweight Detection of Blackhole Attacks in IoT Systems

Abstract

This research introduces a new framework utilizing logistic regression for categorizing nodes within Internet of Things (IoT) networks as either trustworthy or malicious (blackhole). The proposed model integrates various operational characteristics of nodes, such as their success rate in delivering packets, instances of packet loss, energy consumption efficiency, speed of response, collaborative behavior, and dependability. These attributes are mathematically combined through model coefficients and subsequently passed through a sigmoid function to generate a probability score ranging from 0 to 1. A predetermined threshold is then applied to classify nodes; those with a probability above this threshold are deemed trusted, while others are classified as blackhole nodes. A comprehensive mathematical illustration is included to clarify the model's application. To ascertain its efficacy, the framework is subjected to evaluation within simulated IoT network environments. Its performance is subsequently assessed using critical indicators including packet delivery ratio, packet loss rates, latency from source to destination, precision in detection, and the computational burden of routing. The findings indicate that the model effectively identifies compromised nodes, thereby improving the overall robustness and security of the network. This presents a practical and efficient method for real-time threat identification in IoT settings where computational resources are limited.