Deep learning model excels in 5G satellite channel estimation tasks
HELENA for 5G NR LEO NTN Channel Estimation: A Comparative Evaluation
Machine LearningNetworking and Internet ArchitecturePerformance
Summary
Estimating communication channels for satellites in low Earth orbit is tricky due to fast motion and timing issues. The authors checked if a deep learning method designed for ground networks still works well for these satellites once retrained. They found that this method, called HELENA, performs best on accuracy and speed compared to other models on tested datasets. While it works well on powerful hardware, running it quickly enough on smaller devices is still a challenge.
What this means in practice
- •For telecommunications engineers: Improve channel estimation accuracy and speed for satellite 5G communication systems in low Earth orbit using a retrained terrestrial deep learning model.
- •For embedded systems developers: Optimize deployment of deep learning channel estimators on power-constrained hardware for non-terrestrial 5G networks while understanding latency limitations.
Authors
Miguel Camelo Botero, Nina Slamnik-Kriještorac, Johann Marquez-Barja
Abstract
Deep Learning (DL)-based channel estimation has shown high accuracy and low latency in terrestrial 5G NR, but Low Earth Orbit (LEO) Non-Terrestrial Networks (NTNs) introduce Doppler and synchronization impairments that may require NTN-specific architectures. We test whether High-Efficiency Learning-based channel Estimation using dual Neural Attention (HELENA), originally designed for terrestrial channels, remains effective after NTN retraining and suitable across high-performance and power-constrained inference platforms. Its unchanged architecture is trained on paired receiver-compensated (NTN-1) and residual-impaired (NTN-2) datasets and compared with eight terrestrial-origin models trained on the same NTN data and the NTN-specific MDELAN-SISO. HELENA achieves the lowest observed SNR-averaged NMSE among the DL estimators in both conditions, including 55.8-62.7% lower linear-scale NMSE than MDELAN-SISO. All DL models degrade in NTN-2, demonstrating the challenge posed by residual Doppler and its associated impairments. On an RTX PRO 4500, HELENA achieves 0.0595 ms 99th-percentile (P99) inference latency, 88.1% below the 0.5 ms budget, with lower energy than its closest attention-based competitors. On a 10 W Jetson Orin NX, it retains a favorable accuracy-energy trade-off, but no model meets the P99 budget. Thus, HELENA needs no NTN-specific redesign for the evaluated task, while embedded tail latency remains an open challenge.