Unifying RNN and KNN for Enhancing Mirai Attack Detection in IoT Networks

Malicious software gets into Internet of Things (IoT) devices by using exploits and default passwords, which leads to Mirai attacks on IoT networks. These viruses can turn devices into botnets, which raises the risk of major security holes. Massive Distributed Denial-of-Service (DDoS) attacks are then possible, which threatens the security of networks and stops internet services. In This research paper suggests a way to better find Mirai threats in IoT networks by mixing the K-Nearest Neighbors (KNN) and Recurrent Neural Network (RNN) methods. In the proposed model takes the best parts of both RNN and KNN systems and puts them together. The RNN architecture is best at finding time relationships in data about network traffic, while the KNN architecture is best at finding trends. When the authors combined the results of both models into our combined approach, it makes identification more accurate than other methods that are already out there. Empirical studies using a variety of datasets clearly demonstrate that our method regularly and accurately finds Mirai attacks. The Recurrent Neural Network and K-Nearest Neighbors method works better together to protect against new Mirai cyber risks in IoT settings. The integrated RNN and KNN provide the good accuracy with 82% of mirai attack detection. This new development makes a big difference in the field of defense.