BoT-IoT based Denial of Service Detection with Deep Learning

With a sharp rise in the number of internet connected devices, the internet has turned into a necessity for human life. Everyday human life and activities are becoming dependent on Internet of Things (IoT) devices in particular. However, the challenges faced by these IoT devices in terms of data security have been increasing with no well-defined solutions. With the increasing security challenges, several techniques are developed to ensure safe data transfer over IoT networks. Despite these efforts, data breach is observed on a regular basis. The use of machine learning schemes is an optimal solution to enhance the data security in IoT networks. Several deep-learning and machine-learning schemes are discussed in this paper along with standard datasets that may be used for enhancing the IoT network security performance. A deep learning algorithm-based model is developed in this paper for detection of denial-of-service (DoS) attacks. Seaborn, TensorFlow, and scikit-learn python packages are used for this purpose. Efficient attack mitigation and improved accuracy is observed using the proposed model on IoT networks.