![[2602.15326] SCENE OTA-FD: Self-Centering Noncoherent Estimator for Over-the-Air Federated Distillation](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.15326] SCENE OTA-FD: Self-Centering Noncoherent Estimator for Over-the-Air Federated Distillation
Summary
The paper presents SCENE, a novel estimator for over-the-air federated distillation that enhances aggregation without requiring pilot signals, optimizing performance in hardware-constrained environments.
Why It Matters
This research addresses a critical challenge in federated learning by proposing a method that reduces the need for pilot signals, thereby improving efficiency in resource-limited settings. It has implications for the scalability and practicality of federated learning systems, especially in real-world applications where bandwidth and power are constrained.
Key Takeaways
- SCENE eliminates the need for pilot signals in federated distillation.
- The method provides unbiased aggregation while maintaining low variance.
- It is particularly effective in short-coherence and hardware-constrained scenarios.
- The proposed estimator can outperform traditional coherent designs under certain conditions.
- The research contributes to the efficiency of distributed learning systems.
Electrical Engineering and Systems Science > Signal Processing arXiv:2602.15326 (eess) [Submitted on 17 Feb 2026] Title:SCENE OTA-FD: Self-Centering Noncoherent Estimator for Over-the-Air Federated Distillation Authors:Hao Chen, Zavareh Bozorgasl View a PDF of the paper titled SCENE OTA-FD: Self-Centering Noncoherent Estimator for Over-the-Air Federated Distillation, by Hao Chen and Zavareh Bozorgasl View PDF HTML (experimental) Abstract:We propose SCENE (Self-Centering Noncoherent Estimator), a pilot-free and phase-invariant aggregation primitive for over-the-air federated distillation (OTA-FD). Each device maps its soft-label (class-probability) vector to nonnegative transmit energies under constant per-round power and constant-envelope signaling (PAPR near 1). At the server, a self-centering energy estimator removes the noise-energy offset and yields an unbiased estimate of the weighted soft-label average, with variance decaying on the order of 1/(SM) in the number of receive antennas M and repetition factor S. We also develop a pilot-free ratio-normalized variant that cancels unknown large-scale gains, provide a convergence bound consistent with coherent OTA-FD analyses, and present an overhead-based crossover comparison. SCENE targets short-coherence and hardware-constrained regimes, where avoiding per-round CSI is essential: it trades a modest noncoherent variance constant for zero uplink pilots, unbiased aggregation, and hardware-friendly transmission, and can outpe...