The ROC team is hosting talk by Mahdi Sharara, Research Scientist at the CTO Office of VIAVI Solutions, France.
When: January 20, 2026 from 14h00 to 16h00
Where: Room <TBA> at CNAM, 2 rue Conté, 75003 Paris, France
Speaker: Mahdi Sharara, VIAVI Solutions, France
Title: VIAVI AI RAN Scenario Generator (AI-RSG): A Digital-Twin Enabler of Deep and Reinforcement Learning
Abstract: Future 6G networks will rely on AI-driven optimization, yet developing robust AI models faces a fundamental challenge: Insufficient training data; collecting data from operational networks is expensive and risky. In addition, testing AI-based models on live systems can degrade service performance. Digital twin technology addresses these limitations by replicating network behavior in controlled laboratory environments. This presentation introduces VIAVI AI RAN Scenario Generator (AI RSG); a digital twin platform enabling scalable generation of realistic training data for 6G development. The platform simulates thousands of UEs and cells with realistic mobility patterns, radio propagation characteristics, and resource consumption dynamics. Real-world geographical and network topology data can be imported to enhance simulation realism. The AI-RSG enables both Deep Learning and Reinforcement Learning. For supervised learning, the platform generates massive realistic datasets for training predictive models. For reinforcement learning, it provides a risk-free training and validation environment where agents iteratively refine network management policies without operational consequences. The AI-RSG support Open RAN architecture via standardized interfaces, enableing xApp and rApp development and validation prior to live deployment. The presentation will detail the AI-RSG’s architectural capabilities, examine use cases, and discuss how it facilitates the development and deployment of Deep and Reinforcement Learning.
Bio: Mahdi Sharara is a Research Scientist at the CTO Office of VIAVI Solutions, France, working on the intersection of 6G and AI. He focuses on digital twin methodologies, AI-driven network optimization, and machine learning applications for next-generation wireless systems. He received the Diploma degree in Electrical and Telecommunications Engineering from the Lebanese University, Beirut, Lebanon, in 2018, and the M.S. degree in Telecom and Networks from the Lebanese University in collaboration with Saint Joseph University in the same year. He earned his Ph.D. in Telecommunications from Université Paris-Saclay, France, in 2023. He held postdoctoral researcher positions at CentraleSupélec (2023), Orange (2024), and Cergy University (2025) before joining VIAVI Solutions. His research interests include resource allocation in mobile networks, RAN Optimization, deep and reinforcement learning-based algorithms, and Intent Driven networks.
