نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشگاه صنعتی مالک اشتر
2 دانشگاه صنعتی مالک اشت
چکیده .
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
With the rapid growth of satellite communications across various domains, particularly in military applications, and the critical importance of communication security, the use of artificial intelligence-based methods for preventing signal eavesdropping and mitigating interference has gained significant attention. This paper proposes a multilayer perceptron (MLP) neural network-based method for detecting and identifying interference in the DVB-S2 satellite communication standard. The dataset used in this study includes modulated signals (QPSK and APSK with orders of 8, 16, and 32) as well as various types of radio frequency interference, including Chirp Interference (CI), Continuous Wave Interference (CWI), and Multi-Continuous Wave Interference (MCWI). In the proposed approach, a large number of statistical and spectral features are first extracted from the signals. Then, the top ten features are selected using Mahalanobis distance, Euclidean distance, Chi-square, and Variance-to-Mean ratio (VdM) criteria to optimize the model’s performance. Numerical results demonstrate that the VdM criterion improves interference detection accuracy, especially under low signal-to-noise ratio (SNR) conditions. Specifically, at SNR = 0 dB, the proposed method provides up to 6% higher accuracy compared to previous approaches, and at SNR = 9 dB, its accuracy reaches 99.74%.
کلیدواژهها [English]
)