An Iterative Convolutional Neural Network Based Malicious Node Detection (ICNNMND) Protocol For Internet of Things

At present Internet of Things is widely used in various products and parts in our daily life. In the future, its uses and requirements will increase. However, there are several issues with the IoT network, one of which is secure data transmission. Also, malicious nodes in the IoT network are very difficult to identify accurately and quickly, whereas it is very important to do so. In the current context, several solutions are being proposed as there is a need to focus on setting up secure IoT frameworks, and as well as the protocols that runs on them. This research proposes a secure protocol called iterative Convolutional Neural Network based Malicious Node Detection protocol that uses the erroneous assumption learning problematic technique. This protocol will make changes in the application layer, the network layer and the physical layer in order to achieve secure data transmission in IoT. The proposed ICNNMND protocol will be able to detect attacks and eliminate them hence providing maximum data accuracy in IoT. For this, machine learning method will be used and the changes in the characteristics of the malicious nodes will be learned and their boundaries be explored. Then its error variation will be calculated and the malicious nodes be differentiated.