Enhancing smart grid security: A novel approach for efficient attack detection using SMART framework

Abstract

Smart Grids are constructed using control, networking, and advanced computing technologies. Malicious attacks are still a possibility for the grid, although there are now insufficiently fast and accurate detection methods. Cybercrimes affecting the security of the smart grid include the compromise of vital client data by attackers, the spread of viruses, cybersecurity mistakes, and vulnerabilities in distributed systems. In order to provide security in a smart grid context, a SMART (Smart Attack detectoR Technique) framework is developed in this paper. The attacker first sends the request to the risk identification, which will identify the request's original source. Risk estimation calculates the degree of risk and gives processing decisions. Following that, word embedding is used to complete the pre-processing. Information will be prohibited if the process is found to be too vulnerable according to CNN-BiLSTM categorization criteria. The data will be sent to the smart grid automatically if the procedure is low vulnerable and there is no assault. A MATLAB simulator is used to implement the suggested approach. In comparison to the presently used CNN-LSTM [10], SCADA [13], and GOOSE [14] approaches, which have success rates of 19.09 %, 32.22 %, and 57.47 %, respectively, the experimental findings indicate that the suggested SMART strategy has a 99 % success rate, which is very high.

Keywords

• Word embedding
• CNN-BiLSTM
• Smart grid
• SMART
• Network