Mitigating DDoS and DoS attacks in IoT Networks with LSTM-based intrusion detection system
Geetika Singh, Vanshita, Mehakeet Kaur, Pratiksha Goswami · 2025
The number of IoT devices proliferating across the globe is estimated to grow at a rate of 13% by the close of 2024. The IoT technology has numerous applications which include: industrial IoT, smart homes, and smart cities where devices containing sensors and actuators collect and interact with the data(temperature, humidity, sound). The collected data goes through heavy processing and analysis. However, rapid deployment of IoT devices comes with a significant challenge on security. Data breaches, unauthorized access, and services disruption are some of the potential attacks on IoT systems. Traditional method struggles while detecting real-time threats in the network.This paper conducts a comparative study of LSTM-based IDS model for botnet detection in IoT networks. In contrast to previous approaches that reached 97% accuracy, proposed LSTM model demonstrates superior performance with 99% accuracy on the Bot-IoT dataset. In conclusion, as IoT continues to integrate into every sphere of our lives, security becomes paramount. In this paper, development an intrusion detection system(IDS) for Bot-IoT attacks, specifically focusing on distributed denial of service (DDoS) and denial of service (DoS) attacks.