Abstract

Efficient satellite communication is critical for ensuring seamless data transmission across various applications, including remote sensing, defense, and global connectivity. Traditional signal processing techniques face challenges such as signal degradation, interference, and bandwidth limitations, reducing overall transmission efficiency. Advanced optimization algorithms can enhance signal integrity, mitigate noise, and improve data throughput. This study proposes an adaptive hybrid optimization framework integrating Deep Learning- based Channel Estimation (DL-CE) with an Enhanced Error Correction Model (EECM). The DL-CE employs a Convolutional Neural Network (CNN) combined with a Recurrent Neural Network (RNN) to predict channel variations dynamically, reducing transmission errors by 32.5%. Meanwhile, the EECM incorporates Low-Density Parity-Check (LDPC) codes optimized using a Genetic Algorithm (GA) to enhance error correction efficiency, leading to a 27.8% reduction in bit error rate (BER) compared to conventional LDPC codes. Experimental evaluations on real-time satellite transmission datasets demonstrate a 21.3% improvement in spectral efficiency and a 36.4% enhancement in data throughput. Comparative analysis with traditional Orthogonal Frequency-Division Multiplexing (OFDM) and Turbo coding-based error correction confirms that the proposed method achieves a lower BER of 1.02 × 10?³, higher peak signal-to-noise ratio (PSNR) of 42.8 dB, and increased data transmission speed of 1.8 Gbps.

Authors

I. Gugan¹, V. Praveena², Seyed M. Buhari³
Dr. N.G.P. Institute of Technology, India^1,2, Universiti Teknologi Brunei, Brunei³

Keywords

Satellite Communication, Deep Learning, Error Correction, Spectral Efficiency, Data Throughput