Quantum Information Learning

ended

This project aims at developing efficient, scalable and near-term-implementable decoders for long-distance communication on free-space or optical-fiber quantum channels. It is a timely proposal for the establishment of high-data-rate space communication and secure fiber communication. Its importance is recognized in the strategic quantum communication pillar of the EU Flagship on Quantum Technologies. In this setting, the main obstacle towards implementations is a joint quantum measurement, impossible with current technology. This project will investigate how to build a complex decoder by combining simple optical gates and photodetectors, augmented by a classical neural network that controls the quantum parameters and post-processes the measurement data. Moreover, it investigates the power of machine learning for quantum technologies, a very recent interdisciplinary field that still lacks killer applications. The applicant has a PhD and several excellent publications on the subject, while the host group comprises world-leading experts in quantum information theory and pioneers in the application of statistical-inference methods to it. Hence he is in the best position to carry out the project and become an established researcher, who will be acknowledged for his inter-disciplinary expertise and his contribution to an outstanding technological problem. The scientific part of the project is complemented with a well-balanced training program: advanced training-through-research, participation to a machine-learning school, transferable-skills for career development. The researcher will engage in teaching to undergraduate students, the organization of seminars and of a machine-learning reading club. Potential exploitation of the results will take place at a quantum industry day and hackathon. Moreover, outreach events will include an original proposal for the realization of a tape-art exhibition explaining quantum-mechanics by representing quantum-technological protocols.