Security Analysis of IoT Networks using Diffusion Kernel Spike Attention Network with Atomic Optimization based Intrusion Detection System

Recently, the most prominent study areas are Internet of Things (IoT) and its applications. The attributes of the Internet of Things (IoT) facilitate its practical uses in real life, but also leave it vulnerable to online attacks. Numerous IoT nodes gather, store, and handle personal data, making them a prime target for hackers. The main objective of a network intrusion detection system is to find network intrusions, which is essential to guaranteeing the Internet of Things' security. Numerous investigators have tackled this issue and have put out numerous intrusion detection methods. Cyberattack subcategories are hard to identify and performance improvement is a challenge for the existing system. This research project introduces a new method for security analysis in Internet of Things networks by creating the Diffusion Kernel Spike Attention Network with Atomic Optimization (DifK-SAN+AO). This research employs a Diffusion Kernel Spike Attention Network to counteract cyberthreats and provides an overall threat analysis of the Internet of Things. The classification of threat and normal patterns on an IoT network is the main topic of this research study. For every form of attack, the proposed model shows promising findings. We use empirically based, simulation-generated datasets (ToN-IoT, BoT-IoT, and CIC-IoT-23 datasets) to detect attacks directed on the IoT ecosystem. In order to perform a security analysis, the results can be compared using the security metrics that are calculated, including the accuracy, precision, recall, and F1-score factors. The experimental results are more reliable than current algorithms and show 99.2% accuracy in detecting a variety of threats. As it would be observed from the results of the simulations, this demonstrates how efficient our method is in assessing the impacts of the attacks in IoT networks and maintaining good performance values.