Abstract

The rapid evolution of 6G low Earth orbit (LEO) satellite networks presents new challenges in ensuring secure and efficient signal processing at the physical layer. The integration of massive connectivity, dynamic channel variations, and potential eavesdropping threats necessitates robust security mechanisms. Traditional channel estimation techniques struggle to adapt to the highly dynamic nature of LEO satellite channels, leading to degraded performance in secure communications. To address these challenges, a Secure Channel Estimation Model (SCEM) is proposed, leveraging Channel State Information (CSI) and Deep Learning (DL) to enhance physical layer security. The SCEM utilizes a hybrid deep neural network combining Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks to predict CSI with high accuracy. The model is trained and optimized using the D-Wave Leap quantum computing environment to enhance computational efficiency. Experimental evaluations demonstrate a significant improvement in security and signal integrity. The proposed SCEM achieves a 24.7% reduction in bit error rate (BER) compared to conventional Kalman-based estimators and enhances signal-to-noise ratio (SNR) by 8.5 dB. Moreover, the model successfully mitigates eavesdropping risks by improving secrecy capacity by 31.2% over baseline methods. These findings highlight the potential of deep learning in securing next-generation wireless and satellite communications.

Authors

S.K. Rajesh¹, Mary P. Varghese², C. Lisa³, P. Rajkumar⁴, Winson Rajaian⁵
Vidya Academy of Science and Technology, India^1,2, Nehru College of Engineering and Research Centre, India^3,4, University of Technology and Applied Sciences, The Sultanate of Oman⁵

Keywords

6G LEO Satellite Networks, Secure Channel Estimation, Channel State Information, Deep Learning, Physical Layer Security