GD4: Graph-based Discrete Denoising Diffusion for MIMO Detection

Qincheng Lu, Sitao Luan, Xiao-Wen Chang · 2026

Read GD4: Graph-based Discrete Denoising Diffusion for MIMO Detection by Qincheng Lu, Sitao Luan, Xiao-Wen Chang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-00423.txt.

Description

Qincheng Lu, Sitao Luan, and Xiao-Wen Chang speed up signal recovery for MIMO wireless systems, where many antennas transmit and receive at once. Finding the best recovery is provably hard, and the hardest practical case is under-determined, with fewer receive than transmit antennas. Recent diffusion-based detectors show promise but need many sampling steps and degrade exactly in that hard case.

Their method, GD4, recasts a graph-based diffusion process that denoises directly in the discrete symbol space, rather than in a relaxed continuous one. That change allows good answers in one or a few evaluations. Under a similar computing budget, GD4 produced better solutions than rival diffusion detectors and respected classical methods, in both under-determined and over-determined systems. The result is a more favorable trade between solution quality and running cost for receiver design.

Catalog record

CREATORQincheng Lu, Sitao Luan, Xiao-Wen Chang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE32 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.00423

Browse the full library · About the Tower of Babel