A Hybrid CNN/RNN Architecture for Malicious DNS Detection in Electric Vehicles and IoT Devices

Krishnendu Das

Abstract

In the context of smart cities, the integration of Electric Vehicles (EVs) presents new opportunities and challenges related to cyber security, particularly regarding domain name system (DNS) threats. EVs, which are an important component of smart city infrastructure, rely on Internet connectivity for various services, including navigation and real-time traffic updates, making them vulnerable to DNS-based cyberattacks. Malicious DNS activities pose a threat to these integrated systems, potentially disrupting communication and services crucial for EV functionality.

Our proposed lightweight DNS detection model is well suited for deployment on embedded devices found within EVs, ensuring that DNS threats can be recognized and neutralized swiftly, thus maintaining the integrity and efficiency of smart city operations. Using a hybrid CNN and RNN architecture, the model processes the sequence of data effectively, offering protection not only against general malware but also against specific DNS threats that can affect EV communications. This improves the overall cyber resilience of smart cities as they incorporate more advanced and interconnected technologies.

Keywords

Malicious DNS Detection; Lightweight Threat Detection; Convolutional Neural Networks (CNN); Bidirectional Long Short-Term Memory (BiLSTM); Embedded IoT Devices; DNS Security; Smart Traffic Control; Automated Vehicle Systems

Corresponding Author

Krishnendu Das, DST-CIMS, Banaras Hindu University, Varanasi, India.

Citation

Das, K. (2025). A Hybrid CNN/RNN Architecture for Malicious DNS Detection in Electric Vehicles and IoT Devices. Cognitive Computing and Extended Realities, 1(1), 01-16.