ARP spoofing detection using coral reefs optimization algorithm and MLP network

The Address Resolution Protocol (ARP) is a crucial network protocol that maps IP addresses to MAC addresses within local area networks (LANs). ARP is inherently susceptible to spoofing attacks, which involve the injection of forged ARP packets to deceive devices and intercept network traffic, despite its pervasive utility. This paper proposes a novel detection framework that utilizes Coral Reefs Optimization (CRO) to facilitate advanced feature selection and a Multi-Layer Perceptron (MLP) neural network to ensure precise classification. The CRO algorithm effectively identifies the most effective ARP and network-level features, thereby optimizing the input space and enhancing the learning capacity of the neural model. The system accomplishes a robust distinction between legitimate and spoofed ARP packets by training the MLP on this refined feature set. The metaheuristic-driven selection strategy is particularly beneficial in dynamic networking scenarios, as it prevents overfitting and enhances generalization. In contrast to conventional signature-based tools, the proposed CRO-MLP framework is capable of adapting to changing attack vectors and network states, ensuring consistent detection in high-traffic and complex environments. Experimental evaluations conducted in a simulated SDN environment have shown that the method surpasses established baselines in terms of real-time response, computational efficiency, and accuracy. These findings substantiate the CRO-MLP.