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

Collection of recent 2026 arXiv papers on mmWave/THz wireless systems, integrated sensing-and-communication (ISAC), antenna/beamforming methods, and sensor-model foundations. Key contributions include: a conditional diffusion generative prior for efficient probabilistic beam alignment (large improvements in Hit@k and reduced training overhead); a distributed optimization-learning framework using graph transformers for terahertz cell-free ISAC; analysis of a time-varying "disco" reconfigurable intelligent surface (DRIS) showing substantial communication degradation but mixed sensing effects; a呼