Enhancing IoT Routing Security: Performance Analysis of Learning Automata-Firefly Algorithm for RPL under Attack

Abstract

The Internet of Things (IoT) relies heavily on the Routing Protocol for Low-Power and Lossy Networks (RPL). Yet, its open design makes it highly vulnerable to routing disruptions, especially the blackhole attack, where a malicious node advertises deceptive routing metrics and subsequently drops all received packets. This work proposes a security-enhanced RPL framework that integrates Learning Automata (LA) and the Firefly Algorithm (FA) to achieve adaptive, optimized routing under adversarial conditions. A lightweight trust-based mechanism is incorporated to detect and isolate blackhole nodes based on packet-forwarding behavior. The hybrid LA–FA model dynamically learns optimal parent nodes, while the trust system penalizes nodes exhibiting abnormal packet dropping. Simulations conducted in the Contiki Cooja environment with varying node densities show that the blackhole attack significantly impairs network performance, resulting in a nearly 40% reduction in Packet Delivery Ratio (PDR), a 35% decrease in throughput, and an increase in energy consumption exceeding 70%. The suggested LA–FA RPL greatly reduces these impacts, achieving +25% increased PDR, +20% enhanced throughput, −15% reduced delay, and −17% reduced energy use compared to the standard RPL. The findings validate that merging bio-inspired optimization with adaptive trust learning provides an effective and significant approach to safeguard resource-limited IoT routing against blackhole attacks.

Keywords