arXiv：cs.LG· Mohamed Kamel, Tamer Nadeem·· 3 小时前
FlowSep-RF：用 Rectified Flow Matching 实现单通道射频信号分离
FlowSep-RF: Rectified Flow Matching for Single-Channel Radio-Frequency Signal Separation
AI 导读
FlowSep-RF 提出一种 Rectified Flow Matching 框架，学习从接收混合信号到目标信号的干净端点传输，融合时域 IQ 采样与频域特征，仅需一次端点速度评估即可恢复分离波形。
正文
Abstract:Single-channel radio-frequency (RF) signal separation is important for interference mitigation when a signal of interest overlaps with a co-channel interferer in time and frequency. Classical matched filtering and LMMSE receivers are limited under structured non-Gaussian interference, while diffusion and score-based methods often require iterative sampling or optimization. We propose FlowSep-RF, a rectified flow matching framework that learns a clean-endpoint transport from the received mixture to the signal of interest. The model combines time-domain IQ samples with frequency-domain features and recovers the separated waveform using one endpoint velocity evaluation. We evaluate FlowSep-RF on QPSK mixtures with OFDM-BPSK, OFDM-QPSK, and RFChallenge CommSignal2 interference against classical, discriminative, and generative baselines. FlowSep-RF achieves the lowest average BER and MSE among evaluated baselines across all three settings with one-step inference.
| Comments: | 5 pages. Preprint. Under review |
| Subjects: | Signal Processing (eess.SP); Machine Learning (cs.LG) |
| Cite as: | arXiv:2610.20269 [eess.SP] |
| (or arXiv:2610.20269v1 [eess.SP] for this version) | |
| https://doi.org/10.48550/arXiv.2610.20269 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mohamed Kamel [view email]
[v1]
Wed, 14 Oct 2026 14:31:10 UTC (1,193 KB)
来源：arXiv：cs.LG · arxiv.org