Doppler & Timing Advance in Communications

• Doppler and Timing Advance are fundamental phenomena where Doppler accounts for frequency shifts from relative motion and Timing Advance compensates for propagation delays.
• They are crucial for optimizing synchronization and channel estimation in terrestrial, satellite, and optical/space-based communication networks.
• Techniques such as delay-Doppler processing, extended Kalman filtering, and FIR filtering enable robust performance under high mobility and diverse propagation conditions.

Doppler and Timing Advance refer to two fundamental phenomena impacting the synchronization, channel estimation, and noise mitigation in both radio and optical/space-based communication systems. Doppler effects arise from relative motion between transmitters and receivers, leading to frequency shifts and phase drifts, while timing advance (TA) denotes the time adjustment—typically commanded to user devices—needed to pre-compensate for round-trip propagation delays and ensure precise symbol/frame alignment at the receiver. The interplay of Doppler shifts and TA estimation is critical in mobile terrestrial wireless, satellite, and interferometric measurement contexts, particularly under high-mobility or non-terrestrial environments.

1. Doppler Phenomena in Communication Systems

The Doppler effect constitutes a frequency shift observed in a signal when there is a nonzero radial velocity component between transmitter and receiver. In mathematical terms, for a transmitter at carrier frequency $f_c$ and relative line-of-sight velocity $v_n(t)$,

$\Delta f_D(t) = -\frac{v_n(t)}{c}f_c$

where $c$ is the propagation speed ($c\approx 3\times 10^8$ m/s for EM waves in vacuum).

In the context of time-delay interferometry (TDI), such as in the TianQin gravitational wave detector, Doppler shifts induce large, low-frequency phase drifts in heterodyne beatnotes. The measured phase accumulation is

$$\phi^d(t) = -\frac{\omega_0}{c} \int^t n \cdot v(t')\,dt'$$

where $\omega_0$ is the nominal angular frequency and $n$ the line-of-sight unit vector (Zheng et al., 2022).

In wireless terrestrial and satellite networks, Doppler effects manifest as frequency spreads and fading, leading to inter-carrier interference in multicarrier systems and impacting timing synchrony and ranging. High-mobility and LEO satellite scenarios are especially affected, necessitating Doppler-resilient waveform design and estimation algorithms (Sinha et al., 2020, Zhu et al., 2024, Balakrishnan et al., 10 Jun 2025).

2. Timing Advance: Definition and Estimation

Timing advance is the deliberate offset applied at the transmitting device to counteract the two-way propagation delay, so that uplink bursts arrive at the receiver (e.g., base station or satellite) within assigned receive windows. Given a distance $d$ between node and receiver,

$\mathrm{TA} = \frac{2d}{c}$

The TA value is usually estimated at the infrastructure end (base station, gateway, or on-ground processing center) based on preamble round-trip measurements or explicit ranging, and is then fed back to the transmitter as a control command.

Estimation of TA must be robust to multi-path delays, Doppler shifts, and, in satellite scenarios, fast-changing slant ranges. High-resolution delay-Doppler domain analysis, cross-correlation peak detection, and extended Kalman filtering are among the key methodological approaches (Sinha et al., 2020, Balakrishnan et al., 10 Jun 2025, Farhang et al., 2024).

3. Delay-Doppler Domain, OTFS, and SC-FDMA Approaches

Orthogonal Time Frequency Space (OTFS) and its SC-FDMA equivalent offer a framework in which signals are modulated and demodulated on a delay-Doppler (DD) grid. Each resource element is indexed by delay and Doppler bins, with spacings: $$\Delta \tau = \frac{1}{M \Delta f}, \quad \Delta \nu = \frac{1}{N T}$$ where $M$ is the number of delay bins (related to available bandwidth $\Delta f$) and $N$ the Doppler bins (related to symbol duration $T$). The 2D symplectic finite Fourier transform (SFFT) ensures that the inherent channel delay and Doppler spreads are mapped nearly diagonally, simplifying equalization and estimation.

In random-access systems, preambles localized in the delay-Doppler plane allow receivers to separately estimate timing (from delay axis peaks) and Doppler (from Doppler axis energy distribution). Notably, in high Doppler scenarios, well-designed OTFS-based random access (RA) preambles exhibit negligible timing estimation bias due to Doppler—Doppler shifts only disperse energy in the $k$ (Doppler) axis; delay estimation (TA) from the $l$-axis remains accurate (Sinha et al., 2020, Farhang et al., 2024).

4. Joint Doppler and Timing Advance Estimation in LEO Satellite Connectivity

LEO satellite networks present unique challenges: satellites move at several km/s, inducing Doppler shifts in the hundreds of kHz at Ka-band, and TA can span tens to over a hundred ms depending on elevation. Robust estimation demands modeling the joint satellite-UE kinematics, propagation delays, and potentially unsynchronized clocks.

Recursive Bayesian estimation via extended Kalman filters (EKF) is effective in modeling the nonlinear system of satellite and UE motion:

• The state vector includes 3D positions and velocities of the satellite and UE.
• The process model includes Newtonian updates, with satellite acceleration driven by Earth’s gravity.
• Measurement updates incorporate slant-range and elevation from received signals.
• The inferred state yields TA and Doppler via:

$$\mathrm{TA}(t) = 2\|p^\ell(t) - p^u(t)\|/c, \qquad \Delta f(t) = f_T \frac{(v^\ell - v^u) \cdot (p^\ell - p^u)}{\|p^\ell - p^u\| c}$$

• Practicalities such as limited visibility (low-elevation cutoffs) and clock drifts (handled by augmenting the state or post-processing TDoA signals) are incorporated (Balakrishnan et al., 10 Jun 2025).

Simulation studies confirm robust TA and Doppler tracking even at high mobility, with the filter maintaining accuracy under realistic noise and visibility constraints.

5. Doppler and Timing Advance Mitigation in Interferometry: TianQin TDI

Space-based laser interferometry, exemplified by TianQin, relies on heterodyne phase measurements across inter-satellite links. The main challenges are:

• The Doppler-induced phase ramp $\phi^d(t)$, which evolves mainly below $10^{-4}$ Hz.
• Inter-satellite light travel times ('timing advances') that vary with relative motion.
• Strong Earth-Moon gravitational perturbations, giving rise to pronounced low-frequency Doppler signatures.

Mitigation is achieved by:

• Applying high-order linear-phase FIR high-pass filtering to remove the low-frequency $\phi^d(t)$ component from raw beatnote phase data before TDI combination (preferred over post-TDI filtering).
• Compensating FIR filter group delay in delay-operator application.
• Achieving secondary noise limited performance with pseudo-ranging errors up to $\mathcal{O}(10)$ m.
• Maintaining the integrity of TDI combinations, which, via their delay-operator structure, already annihilate $\phi^d$ components that are stationary over the TDI bandwidth (provided sufficient ranging accuracy) (Zheng et al., 2022).

Simulation indicates that this preprocessing pipeline ensures gravitational wave signals are recovered with negligible contamination from Doppler effects or ranging errors, relaxing hardware constraints and linearizing downstream processing.

6. Practical Algorithms and Performance in Terrestrial, Satellite, and Optical Systems

The following algorithms exemplify state-of-the-art approaches to joint Doppler and TA handling:

System Type	Doppler Mitigation	Timing Advance Estimation	Key Results
OTFS-based RA (terrestrial)	DD-domain preamble design	$l$-index search on DD grid	TEP $<10^{-3}$ for $B_c=1.08$ MHz, robust under $\nu_{max}\le 1.2$ kHz (Sinha et al., 2020)
LEO 5G/6G Satellite (LEO)	UE-side pre-compensation, EKF;	Cross-correlation in frequency-domain, cyclic	$<$1% missed detection at SNR = –6 dB, typical TA errors $<$10 sample periods (Zhu et al., 2024, Balakrishnan et al., 10 Jun 2025)
TianQin/Space Interferometry	FIR high-pass removal of $\phi^d$	Delay-operator bookkeeping in TDI	Achieves GW-limited noise with $\leq 10$ m pseudo-range (Zheng et al., 2022)
SC-FDMA (terrestrial/DD view)	Implicit via DD grid structure	Delay axis correlation, “first-peak” estimator	Sub-meter TA accuracy for $B_W\geq 5$ MHz, $N \geq 16$ (Farhang et al., 2024)

Practical guidelines include using delay-Doppler domain 'tiling' for multiuser access, guard-time extensions to tolerate residual TA errors, and robust peak-detection thresholding to eliminate multipath bias in TO/TA extraction.

7. Implications, Ongoing Directions, and Design Guidelines

Robust joint estimation and compensation of Doppler and TA are increasingly crucial as systems operate at higher carrier frequencies, support extreme mobility (e.g., dense LEO constellations), or require sub-meter ranging for synchronization and scientific measurements. Modern waveform designs in both terrestrial and non-terrestrial networks—particularly those leveraging DD domain processing—demonstrate order-of-magnitude gains in TA estimation robustness under severe Doppler, outstripping legacy RACH protocols.

In optical interferometry, frequency-domain pre-cleaning of Doppler drifts is now essential for extracting sub-nanoradian phase modulations (as with gravitational wave signals) without ballooning computational requirements or imposing millimeter-level hardware ranging constraints.

Emerging research continues to refine delay-Doppler waveform theory, power allocation, guard interval tailoring, and Bayesian state-space estimation—each aimed at optimizing the tradeoffs among SNR, TA quantization, Doppler spread, and computational complexity in diverse deployment scenarios (Zheng et al., 2022, Sinha et al., 2020, Zhu et al., 2024, Balakrishnan et al., 10 Jun 2025, Farhang et al., 2024).