6G-Bench: Evaluating Semantic Communication and Network-Level Reasoning in AI-Native 6G Systems

#AI #ML #6G #AINative #SemCom #MachineReasoning #AgenticAI #benchmarking #standards #3GPP #IETF #ETSI #ITUT #ORAN #networks #communications #futurenetworks #ieee #ingrwebinar2026

Special Presentation by Mohamed Amine Ferrag (UAE University, UAE)

Hosted by the Future Networks Artificial Intelligence & Machine Learning (AIML) Working Group

Date/Time: Thursday, 16 April 2026 @ 12:00 UTC (12 PM GMT)

Topic:

6G-Bench: Evaluating Semantic Communication and Network-Level Reasoning in AI-Native 6G Systems 

Abstract:

Emerging sixth-generation (6G) networks are envisioned as AI-native systems in which foundation models act as high-level reasoning and coordination layers above standardized network functions. While large language models (LLMs) have demonstrated strong capabilities in isolated wireless and networking tasks, their ability to perform network-level semantic reasoning over intents, policies, trust constraints, and multi-agent coordination remains insufficiently evaluated.

In this talk, I will present 6G-Bench, an open benchmark designed to rigorously assess semantic communication and network-level reasoning in AI-native 6G environments. The benchmark defines a taxonomy of 30 decision-making tasks aligned with ongoing standardization efforts in 3GPP, IETF, ETSI, ITU-T, and the O-RAN Alliance. These tasks are grouped into five capability categories: intent and policy reasoning, network slicing and resource management, trust and security awareness, AI-native networking and agentic control, and distributed intelligence for emerging 6G use cases.

Starting from over 113,000 realistic 6G operational scenarios, we construct 10,000 very-hard, task-conditioned multiple-choice questions that require multi-step quantitative reasoning under uncertainty and worst-case regret minimization. After automated filtering and expert validation, 3,722 high-confidence questions form the final evaluation set.

I will also present a comprehensive evaluation of 22 contemporary foundation models and discuss key insights for deploying AI reasoning layers in future AI-native 6G networks.

Speaker:

Mohamed Amine Ferrag earned his Bachelor’s, Master’s, Ph.D., and Habilitation degrees in Computer Science from Badji Mokhtar—Annaba University, Algeria, in 2008, 2010, 2014, and 2019, respectively. He served as an Associate Professor at Guelma University, Algeria (2014–2022), and as a Senior Researcher at the NAU-Lincoln Joint Research Center for Intelligent Engineering, Nanjing Agricultural University, China (2019–2022). From 2022 to 2024, he was Lead Researcher at the Technology Innovation Institute (TII), Abu Dhabi, where he led AI-driven cybersecurity research initiatives. In 2025, he joined the United Arab Emirates University (UAEU) as an Associate Professor in the Department of Computer and Network Engineering. His research focuses on cybersecurity and AI-native systems, including wireless network security, network coding security, applied cryptography, blockchain, Generative AI, large language models (LLMs), software security, and AI applications in cybersecurity. He has authored over 200 peer-reviewed publications with more than 16,700 citations and an h-index of 61. He has led international collaborative research projects with institutions in the UK, Australia, USA, Canada, and China, and has created four widely used cybersecurity datasets — Edge-IIoT, FormAI, CyberMetric, and DIA — now extensively adopted by the AI research community. His work has received multiple prestigious awards, including the 2021 IEEE TEM Best Paper Award, the 2022 Scopus Algeria Award, the 2024 ICT Express Best Paper Award, and the 2024 IEEE ComSoc CSIM TC Best Journal Paper Award. He has been listed among Stanford University’s Top 2% Scientists six consecutive times (2020–2025) and was named in the 2025 Clarivate Highly Cited Researchers list. He currently serves as Associate Editor for the IEEE Internet of Things Journal and ICT Express (Elsevier) and is a Senior Member of IEEE.

Brochure (PDF): Webinar-AIML-2026-04-16-Ferrag-6GBench-Brochure.pdf