PhD project title: Complex nanophotonics for a quantum internet
Outline description, including interdisciplinary, intersectoral and international dimensions
Fiber-optics communication, one of the most prominent architectures for (tele-)communications, is widely used to transfer telephone, internet, and cable-TV signals. Despite the high rate/long-haul range of information transmission allowed by such technology, it is well established that the use of quantum light would open new, so far unavailable opportunities for the transmission and manipulation of information. This avenue is currently pursued by academia and companies alike (Google, Microsoft, IBM) towards the establishment of a quantum internet using optical-fibre channels to
sustain and transmit quantum light. However, conventional integrated-optics circuits are very rigid and lack of reconfigurability (different goals require the realisation of specific circuits, which are then hardly adaptable to other tasks), making them unsuited for a full-fledged
In this project, we will explore an entirely new approach to light manipulation and control based on the use of disordered light-scattering media. Such complex media scrambles light, mixing optical many degrees of freedom (direction, phase and polarisation) and providing a powerful method for the manipulation of optically encoded quantum information. Nanophotonic media hold the promise to achieve unprecedented mode-scrambling capabilities due to their reconfigurability and controllability enabled by the use of sophisticated theoretic techniques based on machine learning and compressed sensing. The original application of such novel techniques to quantum optics problems is at the basis of this project, to harness the mode mixing power of disordered optical systems. This project meets perfectly the criteria of interdisciplinarity laid out by CITI-GENS as it combines fundamental theoretic
know-how on machine learning for quantum physics developed within the Quantum Technology group at Queen’s with state-of-the-art experimental capabilities in quantum engineering and manipulation of complex nanophotonic media available in the group led by Prof. S. Gigan at Sorbonne University and LightOn SAS. Such unique expertise is complemented by the facilities that will be made available by the company LightOn, with which the team and Fellow will collaborate to translate the proposal stemming from this project into tangible prototypes.
Quantum information processing and technologies
Quantum machine learning
Fit to CITI-GENS theme(s)
• Information Technology
• Advanced Manufacturing
The proposal is deeply rooted into the Information Technology domain, as it tackles ground-breaking questions for the
entirely new management of information through encoding, manipulation and communication assisted and enhanced by
quantum mechanics. It also addresses the Advanced Manifacturing theme as it will tackle the engineering of complex
nanophotonics media for optimal light-scattering tasks that are at the basis of the architectures that will be explored in
First Supervisor: M Paternostro School: Mathematics and Physics
Second Supervisor: A Ferraro School: Mathematics and Physics
Third Supervisor: S Gigan Company: LightOn
Name of non-HEI partner(s) LightOn SAS
Contribution of non-HEI partner(s) to the project:
LightOn will support the Fellow by:
• providing adequate supervision, mentoring, training, career development opportunities, training in transferable skills;
• offering the Fellow the possibility to develop collaborations with the non-academic sector by hosting them for
determined periods of time during the project;
• providing specific training sessions at Queen’s University Belfast as part of the development of the project.
LightOn will cover the consumables costs associated with the research activities undertaken by the ESR during their time
at their labs and provide access to necessary equipment and infrastructure. We envisage long-term secondments/placements, in particular after the first year of training of the Fellow.
LightOn is a company created in June 2016 as a spinoff of three prestigious laboratories from Paris, the Langevin Institute, the Kastler-Brossel Laboratory, and the statistical physics Laboratory of the Ecole Normale Supérieure in Paris. The company’s goal is to develop novel optical devices for machine learning, based on the research developed in the academic world. The aim is to perform HP machine learning computations, exploiting light propagation in complex systems, using spatial light modulators and cameras. This system is by design relatively close to optical networks exploited in quantum computing.
The company will explore the applicability of their technology for quantum state engineering, both from experimental considerations and from a theoretical standpoint. In particular, LightOn is currently building a cloud-computing platform for non-standard hardware accelerators for machine learning. A great part of this expertise could be used as well for bringing quantum computing in contact with machine learning experts, without the hurdles of new hardware. As such, this project fits fully the company’s medium-to-long term R&D goals. Collaboration with the CITI-GENS Programme will provide LightOn with the opportunity to establish and develop links with Queen’s.
Subject area: Physics, Artificial Intelligence