English

Diffusion-Based Generative Priors for Efficient Beam Alignment in Directional Networks

Signal Processing 2026-04-14 v1 Artificial Intelligence

Abstract

Beam alignment is a key challenge in directional mmWave and THz systems, where narrow beams require accurate yet low-overhead training. Existing learning-based approaches typically predict a single beam and do not quantify uncertainty, limiting adaptive beam sweeping. We recast beam alignment as a generative task and propose a conditional diffusion model that learns a probabilistic beam prior from compact geometric and multipath features. The learned priors guide top-kk sweeps and capture the SNR loss induced by limited probing. Using a ray-traced DeepMIMO scenario with an 8-beam DFT codebook, our best conditional diffusion model achieves strong ranking performance (Hit@1 0.61\approx 0.61, Hit@3 0.90\approx 0.90, Hit@5 0.97\approx 0.97) while preserving SNR at small sweep budgets. Compared with a deterministic classifier baseline, diffusion improves Hit@1 by about 180\%. Results further highlight the importance of informative conditioning and the ability of diffusion sampling to flexibly trade accuracy for computational efficiency. The proposed diffusion framework achieves substantial improvements in small-kk Hit rates, translating into reduced beam training overhead and enabling low-latency, energy-efficient beam alignment for mmWave and THz systems while preserving received SNR.

Keywords

Cite

@article{arxiv.2604.09653,
  title  = {Diffusion-Based Generative Priors for Efficient Beam Alignment in Directional Networks},
  author = {Esraa Fahmy Othman and Lina Bariah and Merouane Debbah},
  journal= {arXiv preprint arXiv:2604.09653},
  year   = {2026}
}
R2 v1 2026-07-01T12:03:26.233Z