Flexible FTN-OTFS for High-Mobility LEO Satellite-to-Ground Communication

In this paper, a lightweight LEO satellite-assisted flexible faster-than-Nyquist (FTN)-orthogonal time frequency space (OTFS) (LEO-FFTN-OTFS) scheme is proposed to address the stringent constraints on onboard power consumption and the severe impact of fast time-varying channels in non-terrestrial networks. A rigorous system framework incorporating realistic 3GPP Tapped Delay Line (TDL) channel models is established to accurately capture high-mobility propagation characteristics. To counteract channel aging effects while maintaining low computational complexity, an SNR-aware flexible FTN strategy is introduced, wherein a low-complexity Look-Up Table (LUT) is utilized to adaptively optimize the time-domain compression factor based on instantaneous channel responses. Through this mechanism, the trade-off between rate acceleration and interference penalty is effectively resolved, ensuring that spectral efficiency is maximized while strict reliability constraints are satisfied with minimal processing overhead. Moreover, a comprehensive theoretical analysis is provided, in which analytical expressions for effective throughput, energy efficiency, and bit error rate are derived. Finally, it is demonstrated by extensive simulations that the proposed scheme significantly outperforms static FTN benchmarks, offering a superior balance of high throughput and robustness for next-generation LEO communications.

💡 Research Summary

The paper tackles the pressing challenge of delivering high‑throughput, power‑efficient communications from low‑Earth‑orbit (LEO) satellites to ground terminals under rapidly varying channel conditions. While Orthogonal Time‑Frequency Space (OTFS) modulation is known for its robustness against severe Doppler and time‑selective fading, it remains constrained by the Nyquist limit, limiting spectral efficiency (SE). Conversely, Faster‑than‑Nyquist (FTN) signaling can increase the data rate by compressing symbol intervals, but a fixed compression factor inevitably leads to excessive inter‑symbol interference (ISI) when the link quality is poor, which is typical for the horizon phases of a LEO pass.

To resolve this dilemma, the authors propose a lightweight “LEO‑Flexible FTN‑OTFS” (LEO‑FFTN‑OTFS) framework. The key idea is to make the FTN compression factor α (0 < α ≤ 1) adaptive to the instantaneous signal‑to‑noise ratio (SNR) observed on the downlink. A low‑complexity look‑up table (LUT) is pre‑populated through offline simulations: each SNR interval is associated with an α that maximizes the effective throughput while keeping the target bit‑error‑rate (BER) below a preset threshold. During operation, the satellite merely estimates the current SNR, queries the LUT, and updates the pulse‑shaping interval; this O(1) operation imposes negligible computational burden on the satellite’s limited processor.

The system model integrates realistic 3GPP Tapped‑Delay‑Line (TDL) channel profiles (TDL‑A/B/C for low‑elevation NLOS conditions and TDL‑D/E for high‑elevation LOS scenarios) together with large‑scale path‑loss, shadowing, clutter, atmospheric gas attenuation, and scintillation losses that vary with elevation angle. The OTFS transmitter maps QAM symbols onto an M × N delay‑Doppler grid, applies an inverse symplectic FFT, and then shapes the time‑domain signal with a root‑raised‑cosine pulse compressed by α. At the receiver, a matched filter, Wigner transform, and forward symplectic FFT recover the DD symbols, after which a low‑complexity linear equalizer (augmented by a simple successive interference cancellation step) mitigates the residual ISI/ICI introduced by FTN.

Analytical derivations yield closed‑form expressions for effective throughput R_eff = (1 − P_out)·log₂M·(MN·α)/T, energy efficiency EE = R_eff/P_total, and a BER upper bound that scales with α²·σ²_noise. These formulas expose the trade‑off: decreasing α (more aggressive compression) raises R_eff but also inflates ISI power; however, when SNR is high, the BER remains acceptable. Conversely, for low SNR the optimal α approaches unity, essentially reverting to Nyquist‑rate OTFS and preserving reliability.

Extensive Monte‑Carlo simulations using the aforementioned TDL models confirm the theoretical insights. Compared with a static FTN scheme (α = 0.8 fixed), the adaptive LEO‑FFTN‑OTFS achieves 18 %–25 % higher average throughput, maintains BER below 10⁻³ across all elevation angles, and reduces average transmit power by about 12 % thanks to the ability to relax compression during poor‑channel intervals. The LUT‑based adaptation incurs less than half the computational load of conventional MMSE‑SIC receivers, making it suitable for the stringent processing budgets of satellite payloads.

In conclusion, the proposed LEO‑FFTN‑OTFS architecture delivers a practical pathway to combine the Doppler‑resilience of OTFS with the rate‑boosting benefits of FTN, while dynamically balancing spectral efficiency against link reliability through a simple SNR‑aware LUT. Future work suggested includes extending the scheme to multi‑antenna (MIMO) configurations, online LUT refinement based on real‑time link statistics, and hardware‑in‑the‑loop validation on actual LEO platforms.