Tag
Proposes CRFCAN, a complex-valued residual network for joint channel and phase noise estimation in sub-THz OFDM systems, which achieves end-to-end recovery with physics-inspired structure and outperforms conventional and state-of-the-art deep learning models.