Detection of Various Security Attacks on IoT Devices Using Multi-Layer Neural Network Model Over Sensor Networks

In a traditional network environment firewalls are used to filter user access, IoT network environment is not like a client server interaction environment model, firewalls are not possible to deploy into IoT devices, and attackers easily force attack on IoT device to hack valuable information. In this paper we proposed multi-layer neural network model to detect attacks on IoT devices. Neural network formed using set of neurons, each neuron is capable to receive signals from other neurons, at each neuron inputs are received and multiplied with weight values, and calculated output values are transferred to other neurons. In our proposed neural network model based IoT security model deploy neurons into input layer, hidden layers, and output layer. Neurons in this model are fully connected with each other. Performance evaluation metrics like true positive rate, false positive rate precision and accuracy are used to evaluate performance of proposed model.