Devising Network Intrusion Detection System for Smart City with an Ensemble of Optimization and Deep Learning Techniques

In recent years, there is a great advancement in information and communication technologies such as radical improvement in processor units, memory and communication devices. Incidentally, these rapid technological advancements are used by the smart city applications to manage all the resources. Subsequently, the volume of data being acquired, processed, transferred and stored using these high-end technological devices in the smart city are enormous. Besides, the data must be protected from the cyber-attacks. An approach of identifying an attack and perform the necessary operations when there is an attack is known as intrusion detection system (IDS). Mostly, the attacks take place from external entities. The idea of NIDS is proposed for enhancing the security level of smart city from various external attacks. Lately, artificial intelligence methods have been applied in identifying the attack and usage of deep learning is the contemporary approach. In this research, we suggest a network intrusion detection system by combining optimization algorithm and deep learning techniques to identify any intrusions in smart city. Further, the suggested model utilized the improved golden jackal optimization algorithm for optimization. Afterwards, an artificial neural network (ANN) is employed to train and test the model. The CIC-IDS 2017 dataset is utilized for training and testing the model. The accuracy and recall are the indicators used in this research to test the performance of the model. The train test ratio is 80:20. The proposed model outperformed the benchmark model with mean square error of training and accuracy and recall in testing.