Abstract When multimode optical fibers are perturbed, the data that is transmitted through them is scrambled. This presents a major difficulty for many possible applications, such as multimode fiber based telecommunication and endoscopy. To overcome this challenge, a deep learning approach that generalizes over mechanical perturbations is presented. Using this approach, successful reconstruction of the input images from intensity-only measurements of speckle patterns at the output of a 1.5 m-long randomly perturbed multimode fiber is demonstrated. The model’s success is explained by hidden correlations in the speckle of random fiber conformations.

We propose and experimentally demonstrate high-speed operation of random-channel cryptography (RCC) in multimode fibers. RCC is a key generation and distribution method based on the random channel state of a multimode fiber and multi-dimension to single-dimension projection. The reciprocal intensity transmittance of the channel shared between the two legitimate users is used to generate and distribute correlated keys. In previous work, RCC's key rate-distance product was limited by the speed of light. In this work, we show that adding a fast modulator at one end of the channel decouples the key rate and distance, resulting in a significant improvement in the key rate-distance product, limited only by the fiber's modal dispersion. Error-free transmission at a key rate-distance product of 64.7 Mbps 12 km, which is seven orders of magnitude higher than the previous demonstration, was achieved. The proposed method's security arises from a fundamental asymmetry between the eavesdroppers and legitimate users measurement complexity.