{"ID":2873089,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.07839","arxiv_id":"2509.07839","title":"Enhancements in Score-based Channel Estimation for Real-Time Wireless Systems","abstract":"We propose enhancements to score-based generative modeling techniques for low-latency pilot-based channel estimation in a point-to-point single-carrier multiple-input multiple-output (MIMO) wireless system. Building on recent advances in score-based models, we investigate a specific noise schedule design and sampling acceleration by step-skipping to reduce the number of denoising steps during inference. We additionally propose a single-step signal-to-noise ratio informed denoiser as an extreme case of the step-skipping approach. Our methods achieve significant latency reductions without performance degradation, as demonstrated on a synthetic channel dataset representing an urban macrocell MIMO communications scenario.","short_abstract":"We propose enhancements to score-based generative modeling techniques for low-latency pilot-based channel estimation in a point-to-point single-carrier multiple-input multiple-output (MIMO) wireless system. Building on recent advances in score-based models, we investigate a specific noise schedule design and sampling a...","url_abs":"https://arxiv.org/abs/2509.07839","url_pdf":"https://arxiv.org/pdf/2509.07839v1","authors":"[\"Florian Strasser\",\"Marion Bäro\",\"Wolfgang Utschick\"]","published":"2025-09-09T15:13:15Z","proceeding":"eess.SP","tasks":"[\"eess.SP\"]","methods":"[]","has_code":false}