---
title: "Consistency Learning and Distillation for One-Step MIMO-OFDM Channel Estimation"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.06271v1"
permalink: "https://feed.astralor.com/items/itm_6e21c7177f4e997bef59b77d"
published: "2026-11-04T10:16:11.000Z"
score: 56
channel: "tech"
tags: ["推理效率","多模态","Consistency Learning","Distillation","One-Step MIMO-OFDM Channel","Estimation Iterative","Consistency Mid-Trai","Iterative"]
---

# Consistency Learning and Distillation for One-Step MIMO-OFDM Channel Estimation

## At a glance

Iterative neural channel estimators can improve recovery under pilot contamination, but repeated network evaluations increase inference cost. This paper develops observation-anchored consistency and Consistency Mid-Trai…

## Key points

- Iterative neural channel estimators can improve recovery under pilot contamination, but repeated network evaluations increase inference cost.
- This paper develops observation-anchored consistency and Consistency Mid-Training (CMT) estimators for one-step channel estimation in multiple-input multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM)…
- Both estimators refine pilot-derived least-squares channel estimates in the delay-beam domain.

## Full text

Iterative neural channel estimators can improve recovery under pilot contamination, but repeated network evaluations increase inference cost. This paper develops observation-anchored consistency and Consistency Mid-Training (CMT) estimators for one-step channel estimation in multiple-input multiple-output orthogonal frequency-division multiplexing (MIMO-OFDM) systems. Both estimators refine pilot-derived least-squares channel estimates in the delay-beam domain. Consistency learning couples endpoint predictions along noisy-to-clean paths. CMT then refines a compact consistency-trained student to predict a flow-matching teacher's endpoint from intermediate trajectory states. Across the evaluated SNR range, the compact CMT student reduces uncoded symbol error rate (SER) by 9.6-20.2% relative to the one-step flow-matching baseline (0.44-0.98 dB in logarithmic SER reduction), while using 74.5% fewer parameters and 73.0% fewer network multiply-accumulates.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_6e21c7177f4e997bef59b77d