PhD Thesis in Optical Communication

Organisation/Company Orange

Department Orange INNOV

Research Field Engineering » Communication engineering

Researcher Profile Recognised Researcher (R2)

Positions PhD Positions

Country France

Application Deadline 30 Sep 2025 - 12:00 (Europe/Paris)

Type of Contract Temporary

Job Status Full-time

Is the job funded through the EU Research Framework Programme? Not funded by a EU programme

Is the Job related to staff position within a Research Infrastructure? No

Offer Description

Your role is to conduct a PhD thesis on the optimization of the design (via AI models) of discrete and distributed Raman amplifiers and the associated multi-band optical transmission systems.

To increase the bandwidth of optical fibers, one solution is to exploit new amplification bands (O, E, S, L, U). Compared to erbium-doped fiber amplifiers, Raman amplifiers have the advantage of amplifying the optical signal over spectral ranges other than the C-band. They allow, through the multiplexing of several pump wavelengths, to obtain broader gain spectra. However, stimulated Raman scattering (SRS), which is exploited to achieve the gain, makes it difficult to obtain uniform gain across the entire band due to energy exchanges between bands. To mitigate this effect, one approach is to calibrate and optimize the parameters (number and values of pump wavelengths, pump powers) of the Raman amplifier. However, this optimization is complex, as SRS is modeled by a system of coupled nonlinear differential equations, which lacks an analytical solution and whose resolution through numerical methods is computationally expensive. In this context, a promising approach relies on using AI models to automatically calibrate the amplifier parameters. This approach will be generalized to compensate for the effect of SRS in a multi-band optical transmission line through the adjustment of the parameters of discrete and distributed Raman amplifiers.

The objective of the thesis is to assess the relevance of new AI models for the optimization of Raman amplifier parameters. Classic supervised approaches rely on the exploitation of large datasets, which may not necessarily be suited for finding optimal configurations. In contrast to this "black box" approach, an alternative two-step approach is being considered.

In the first step, the focus will be on modeling SRS using advanced Deep Learning techniques, notably based on Physics-Informed Neural Networks (PINNs), Deep Unfolding, and Parametric Networks. The goal is to develop a model with very low complexity capable of capturing the underlying dynamics of Raman amplification by incorporating physical knowledge of the phenomenon.

In the second step, these models will be used to address the inverse problem, namely the optimization of amplifier parameters. This optimization will be carried out using AI techniques such as Deep Learning and Reinforcement Learning, with appropriate metrics for the problem defined in advance.

The effectiveness of these approaches will be validated both in simulation and on the multi-band optical transmission bench at the Orange INNOV laboratory.

What Makes This Offer Attractive

This subject is part of several ongoing cooperative projects (CELTIC-NEXT EMBRACE, France Relance SIMBADE, MSCA Mentor), during which it was demonstrated:

• How a distributed Raman amplifier with three pump wavelengths calibrated empirically could compensate for the effect of SRS in a 100-km SSMF fiber section for multi-band transmission (S+C+L).
• How PINN-type networks effectively and cost-effectively simulate the effect of SRS in a discrete Raman amplifier.

By introducing an optimized AI-based approach, the scientific and technical contributions of this thesis work should improve system performance and validate it experimentally. This proof of concept could therefore facilitate the rapid deployment of multi-band systems based on Raman amplifiers.

Skills (scientific and technical) and personal qualities required for the position:

• Strong knowledge of digital and optical communications (fiber optic transmission technologies).
• Knowledge of AI, machine learning techniques, and neural networks.
• Solid theoretical knowledge and good abstraction skills.
• Programming skills in Python or Matlab.
• Open-mindedness and curiosity.
• Ability to communicate and write in English in a professional context.

Required educational background:

Master’s or engineering degree in digital communications/photonic systems.

Desired experience (internships, ...):

Ideally, one or more internships in a company or research laboratory on topics related to photonic systems or, more generally, digital communications.

Additional Information

Entity
The ambition of the Innovation Division is to further Orange's innovation and strengthen its technological leadership by mobilizing our research capabilities to promote responsible innovation for the benefit of humanity, inform the Group’s long-term strategic decisions, and influence the global digital ecosystem.

We train experts in today’s and tomorrow’s technologies and ensure continuous improvement of our services’ performance and effectiveness. The Innovation division gathers 6,000 employees worldwide dedicated to research and innovation, including 740 researchers. With a global vision and a wide range of profiles (researchers, engineers, designers, developers, data scientists, sociologists, graphic designers, marketers, cybersecurity experts, etc.), the people of Innovation are attentive and serve the countries, regions, and business units to make Orange a trusted multi-service operator.

Within the NETWORKS/WNI department of Orange INNOVATION, you will be integrated into a research team (AOT/SOAN) at the forefront of innovation and expertise on the networks of the future (800G/1.6T, multi-band transmission, AI for Networks…). You will be part of a research ecosystem combining experienced engineers for research and shorter-term studies, enabling the concrete implementation of the studied concepts, benefiting from experimental platforms in optical transport.

The work will be carried out in collaboration with the CNRS Lab-STICC research units (UMR 6285) and the FOTON Institute (UMR 6082).