Simultaneously Sorting of Fractional Vortex Beams (FVBs) Based on Diffractive Optical Neural Network

To fully leverage the advantages of fractional vortex beams (FVBs), accurate sorting of it is essential. However, traditional OAM sorting methods struggle to effectively handle the challenges posed by non-integer OAM states. In this paper, we propose an efficient sorting method based on a diffractive optical neural network (DONN), which demonstrates excellent performance in sorting FVBs and excels in minimizing energy crosstalk between different channels. To validate the practical effectiveness of our sorting device, we built an experimental platform and achieved results consistent with theoretical predictions. This rapid sorting method, combining optical neural networks, features light-speed operation and high efficiency, holding great potential for future FVB sorting and applications.