# Learning How to Demodulate from Few Pilots via Meta-Learning

arXiv: https://arxiv.org/abs/1903.02184 · HTML: https://mlsanity.com/paper/1903.02184/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 6 Mar 2019.

## Built on by (most cited first, 2 of 2)

- [Adapt under Attack and Domain Shift: Unified Adversarial Meta-Learning and Domain Adaptation for Robust Automatic Modulation Classification](https://mlsanity.com/paper/2511.01172/) (2025)
- [Geometric Inductive Biases for Semi-Supervised Equalization: The Constellation-Aware Transformer](https://mlsanity.com/paper/2609.33695/) (2026)

## Abstract (arXiv)

Consider an Internet-of-Things (IoT) scenario in which devices transmit sporadically using short packets with few pilot symbols. Each device transmits over a fading channel and is characterized by an amplifier with a unique non-linear transfer function. The number of pilots is generally insufficient to obtain an accurate estimate of the end-to-end channel, which includes the effects of fading and of the amplifier's distortion. This paper proposes to tackle this problem using meta-learning. Accordingly, pilots from previous IoT transmissions are used as meta-training in order to learn a demodulator that is able to quickly adapt to new end-to-end channel conditions from few pilots. Numerical results validate the advantages of the approach as compared to training schemes that either do not leverage prior transmissions or apply a standard learning algorithm on previously received data.