Derin Öğrenme Tabanlı Kanal Kestirimi: 5G NR OFDM Sistemleri için Hiperparametre Optimizasyonu ve Performans Karşılaştırması

Öz

Kablosuz haberleşme sistemlerinde kanal kestirimi, özellikle dik frekans bölmeli çoğullama (OFDM) gibi modern teknolojilerde kritik bir rol oynamaktadır. Geleneksel yöntemler olan en küçük kareler (LS) ve en küçük ortalama karesel hata (MMSE) teknikleri yaygın olarak kullanılmaktadır, ancak LS düşük SNR koşullarında performans kaybederken, MMSE'nin yüksek işlem karmaşıklığı gerçek zamanlı uygulamalarda dezavantaj yaratmaktadır. Bu nedenle, daha uyarlanabilir ve verimli tekniklere ihtiyaç duyulmaktadır. Derin öğrenme (DL), kablosuz sinyallerin karmaşık yapısını modelleme yeteneğiyle umut verici bir alternatif olarak öne çıkmıştır. Evrişimli sinir ağları gibi DL modelleri, büyük veri setlerinden öğrenerek dinamik ortamlara uyum sağlayabilir. Bu çalışmada, farklı hiperparametrelerle DL tabanlı kanal kestirimi performansı 5G OFDM sistemlerinde analiz edilmiş ve LS yöntemi ile karşılaştırılmıştır. Sonuçlar, optimum hiperparametrelerle DL modellerinin daha etkili sonuçlar verdiğini göstermektedir.

Anahtar Kelimeler

• Kanal Kestirimi
• OFDM
• Kablosuz Haberleşme
• Derin Öğrenme
• Parametre Optimizasyonu

Deep Learning-Based Channel Estimation: Hyperparameter Optimization and Performance Comparison for 5G NR OFDM Systems

Öz

Channel estimation plays a critical role in wireless communication systems, especially in modern technologies like orthogonal frequency division multiplexing (OFDM). Traditional channel estimation techniques, such as least squares (LS) and minimum mean square error (MMSE), are widely used but have performance limitations due to their inherent characteristics. While LS offers low computational complexity, its performance deteriorates in low signal-to-noise ratio conditions. MMSE, on the other hand, provides more accurate estimates but its high computational complexity is a disadvantage for real-time applications. Therefore, developing more adaptable and efficient methods for wireless systems is essential. Deep learning (DL) has emerged as a promising alternative due to its ability to effectively model the complex nature of wireless signals. Models like convolutional neural networks can learn features from large datasets and adapt to dynamic and nonlinear environments. This study analyzes DL-based channel estimation performance for 5G OFDM systems under different hyperparameters, comparing it to the traditional LS method.

Anahtar Kelimeler

• Channel Estimation
• OFDM
• Deep Learning
• Wireless Communication
• Parameter Optimization

Kaynakça

1. D. Tse and P. Viswanath, Fundamentals of Wireless Communication, Cambridge Univ. Press, Cambridge, U.K., 2005.
2. Y. Li and G.L. Stüber, “Orthogonal Frequency Division Multiplexing for Wireless Communications,” in Signals and Communication Technology, Springer, New York, NY, 2006.
3. T. Rappaport, Wireless Communications Principles and Practice, 2nd ed., Prentice Hall PTR, Upper Saddle River, NJ, 2001.
4. S.M. Kay, Fundamentals of Statistical Signal Processing, Volume I: Estimation Theory, 1st ed., Pearson, Boston, MA, 1993.
5. I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning, MIT Press, Cambridge, MA, 2016.
6. C.K. Wen, W. T. Shih, and S. Jin, “Deep learning for massive MIMO CSI feedback,” IEEE Wirel. Commun. Lett., vol. 7, no. 5, pp. 748–751, Oct. 2018.
7. H. Ye, G.Y. Li, and B.H. Juang, “Power of deep learning for channel estimation and signal detection in OFDM systems,” IEEE Wirel. Commun. Lett., vol. 7, no. 1, pp. 114–117, Feb. 2018.
8. M. Belgiovine, K. Sankhe, C. Bocanegra, D. Roy, and K. R. Chowdhury, “Deep learning at the edge for channel estimation in beyond-5G massive MIMO,” IEEE Wirel. Commun., vol. 28, no. 2, pp. 19–25, Apr. 2021.

9. P. Saikrishna, A.K.R. Chavva, M. Beniwal, and A. Goyal, “Deep learning-based channel estimation with flexible delay and Doppler networks for 5G NR,” in Proc. IEEE Wirel. Commun. Netw. Conf. (WCNC), IEEE, 2021.
10. P.K. Chundi, X. Wang, and M. Seok, “Channel estimation using deep learning on an FPGA for 5G millimeter-wave communication systems,” IEEE Trans. Circuits Syst. I, Reg. Papers, vol. 69, no. 2, pp. 908–918, Feb. 2022.
11. A.S.M. Mohammed, A.I.A. Taman, A.M. Hassan, and A. Zekry, “Deep learning channel estimation for OFDM 5G systems with different channel models,” Wirel. Pers.Commun., vol. 128, no. 4, pp. 2891–2912, Feb. 2023.
12. X. Guo, X. Liu, and X. Gong, “Deep learning based channel estimation for 5G,” in Proc. 6th Int. Conf. Inf. Commun. Signal Process. (ICICSP), IEEE, 2023, pp. 700–705.
13. Y. Singh, P. Swami, V. Bhatia, and P. Brida, “Channel estimation in 5G and beyond networks using deep learning,” in Proc. 34th Int. Conf. Radioelektronika (RADIOELEKTRONIKA), IEEE, 2024.
14. S.S. Haykin, Adaptive Filter Theory, 5th ed., Pearson, Boston, MA, 2014.
15. C.T. Nguyen et al., “Transfer learning for wireless networks: A comprehensive survey,” Proc. IEEE, vol. 110, no. 8, pp. 1073–1115, Aug. 2022.
16. Y. Cheng, D. Wang, P. Zhou, and T. Zhang, “Model compression and acceleration for deep neural networks: The principles, progress, and challenges,” IEEE Signal Process. Mag., vol. 35, no. 1, pp. 126–136, Jan. 2018.
17. TSGR, TS 138 214 - V16.2.0 - 5G; NR; Physical layer procedures for data (3GPP TS 38.214 version 16.2.0 Release 16), 2020. Available: https://portal.etsi.org/TB/ETSIDeliverableStatus.aspx.
18. TSGR, TS 138 212 - V16.2.0 - 5G; NR; Multiplexing and channel coding (3GPP TS 38.212 version 16.2.0 Release 16), 2020. Available: https://portal.etsi.org/TB/ETSIDeliverableStatus.aspx.
19. TSGR, TS 138 211 - V16.2.0 - 5G; NR; Physical channels and modulation (3GPP TS 38.211 version 16.2.0 Release 16), 2020. Available: https://portal.etsi.org/TB/ETSIDeliverableStatus.aspx.
20. TSGR, TR 138 901 - V16.1.0 - 5G; Study on channel model for frequencies from 0.5 to 100 GHz (3GPP TR 38.901 version 16.1.0 Release 16), 2020. Available: https://portal.etsi.org/TB/ETSIDeliverableStatus.aspx.
21. TSGR, TS 138 213 - V16.2.0 - 5G; NR; Physical layer procedures for control (3GPP TS 38.213 version 16.2.0 Release 16), 2020. Available: https://portal.etsi.org/TB/ETSIDeliverableStatus.aspx.