A Physics-Guided Receiver for WiFi that Learns Hardware Nonlinearities without Training Data

Read full text Summarise with VIRAT
Authors: Ramakrishna Rao Pasupuleti
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Modern IEEE 802.11ax receivers must compensate power-amplifier (PA) nonlinearity to support high-order modulation, yet the HE-LTF preamble of a standard 52-tone resource unit provides only four pilot subcarriers. We show that high-dimensional residual learning from four observations is severely rank-deficient and unreliable. We instead propose a single-parameter receiver that estimates the Rapp-model saturation amplitude from the four pilots, inverts the known nonlinearity, and gates the correction using a pilot-domain distortion metric that suppresses intervention under near-linear hardware. We establish local identifiability and vanishing-noise consistency of the estimator and prove two safety propositions characterising the gate. Across n≥3000 Monte Carlo trials per condition, the receiver yields statistically significant spectral-efficiency gains at all tested SNR values, preserves gains across 64-, 256- and 1024-QAM modulation, generalises across four channel models, and outperforms a representative deep-unfolded baseline throughout the practically relevant nonlinear operating regime while remaining conservative under clean hardware.

Fields of Study & Subjects

Advanced Power Amplifier Design PAPR reduction in OFDM Advanced Wireless Communication Techniques

Publication details

Journal / source
Zenodo (CERN European Organization for Nuclear Research)
Publication date
22 March 2026
Language
EN
Document type
Preprint