2024 : 7 : 16
Jamal Ghasemi

Jamal Ghasemi

Academic rank: Associate Professor
Education: PhD.
Faculty: Faculty of Technology and Engineering
Phone: 01135302902


Intrusion detection system for network by using deep learning and feature selection
Internet of Things (IoT), intrusion detection system, classification, self-adaptive genetic algorithm, neural network
Researchers Mustafa Jamal Abdul Satar(Student)، Jamal Ghasemi(PrimaryAdvisor)


Today, with the development of technology and the growth of the Internet of Things, cyber-attacks are on the rise. IoT devices can be exposed to various threats and dangers of hackers and malware, including information theft, forgery and denial of service, and cause great material and moral damage to individuals and organizations. Therefore, it is necessary to take security measures in this regard. Recent research into the security mechanisms of IoT devices and intrusion detection systems shows the widespread use of machine learning (ML) algorithms to detect malicious traffic that uses a neural network to learn a model for sequencing communications. It is between computers in a network, and by analyzing and selecting the right features, congested attacks are more accurately detected. The proposed approach, which is a combination of feature selection methods based on self-adaptive genetic algorithm and deep neural network, works in four different steps. In the first step, by entering the data into the proposed system, the data are divided into two categories of training and testing. In the second step, the data is refined and additional features are removed. The purpose of this step is to monitor the quality of the features and improve the classification power. In the third step, a classification based on deep learning neural networks conducts training using training data. At this stage, neural networks learn by regulating the weight of neurons in different layers. In the fourth step of the proposed method, the experimental data set is entered into the system and we predict the sessions using classification based on deep learning neural networks. The results of the evaluations, which are based on the benchmark criteria used to classify the data, show that the proposed method has a much better result than the comparison method.