AI-ML-in-5G-Challenge.github.io

ITU AI/ML Challenge

The ITU AI/ML Challenge offers carefully curated problem statements, a mix of real-world and simulated data, technical webinars, mentoring, and hands-on sessions. Teams participating in the Challenge enable, create, train and deploy ML models for various domains including communication networks. This enables participants to not only showcase their talent, test their concepts on real data and real-world problems, and compete for global recognition including prize money and certificates, but also enter the world of ITU standards by mapping their solutions to our specifications. Since the start of the Challenge in 2020, we have hosted more than 77 competitions (problem statements). There are several Challenge categories that have been hosted until now including AI/ML in 5G Challenge, GeoAI Challenge, tinyML Challenge and AI for Fusion Energy Challenge.

Challenge themes and problem statements

AI/ML in 5G Challenge

Since its inception in 2020, the ITU AI/ML in 5G Challenge has annually convened a diverse community of innovators, encompassing students and professionals from across the globe, to explore and address the cutting-edge challenges of integrating AI and ML technologies into 5G and emerging 6G networks. Through these competitions, the Challenge has become instrumental in driving forward the integration of advanced technologies in communication networks. Below are the various problem statements hosted since 2020;

2023

• 62.AI/ML for 5G-Energy Consumption Modelling---curated by Huawei

• 61. Depth Map Estimation in 6G mmWave systems---curated by NIST

• 60. Fault Impact Analysis: Towards Service-Oriented Network Operation & Maintenance---curated by Huawei

• 59. Graph Neural Networking Challenge 2023 - Creating a Network Digital Twin with Real Network Data---curated by BNN-UPC

• 58. Intrusion and Vulnerability Detection in Software-Defined Networks (SDN)---curated by ULAK Comm.

• 57. Multi-environment automotive QoS prediction---curated by Fraunhofer HHI

• 56. Network Traffic Scenario Prediction Challenge---curated by ZTE

• 55. QoS Prediction Challenge---curated by Fraunhofer HHI

• 53. Title Extraction in Lecture Slides Challenge---curated by ITU

• 53. Network failure classification model using network digital twin---curated by KDDI

• 52. Multi Modal V2V Beam Prediction Challenge 2023---curated by Wireless Intelligence Lab - Arizona State University

• 51. 3D Location Estimation Using RSSI of Wireless LAN---curated by RISING - JAPAN

• 50. Build-a-thon 2023---curated by ITU Focus Group on Autonomous Networks (FG-AN)

2022

• 50. BYOC: Build your own Closed loop---curated by ITU Focus Group Autonomous Networks (FG-AN)

• 49. Classification of Home Network Users to Improve User Experience---curated by ZTE

• 48. Depth Map Estimation in 6G mmWave systems---curated by NIST

• 47. Federated Traffic Prediction for 5G and Beyond---curated by CTTC (Centre Tecnològic de Telecomunicacions de Catalunya)

• 46. Graph Neural Networking Challenge 2022: Improving Network Digital Twins through Data-centric AI---curated by BNN-UPC

• 45. I/Q-based Beam Classification with the DeepBeam Dataset---curated by Northeastern University

• 44. Location Estimation Using RSSI of Wireless LAN in NLoS Environment---curated by RISING

• 43. Machine Learning for Throughput Prediction in Coordinated IEEE 802.11be Wi-Fi networks---curated by UPF

• 42. Multi Modal Beam Prediction Challenge 2022: Towards Generalization---curated by Arizona State University

• 41. Network failure prediction on CNFs 5GC with Linux eBPF---curated by KDDI

• 39. Next-Gen WiFi Throughput Prediction Challenge---curated by ITU, UPF

• 40. Non-linear Power Amplifier Behavioral Modeling to achieve higher energy efficiency in 5G RAN---curated by ZTE

• 39. "Slidin' videos": Slide Transition Detection and Title Extraction in Lecture Videos---curated by ITU

• 38. Synthetic Observability Data Generation using GANs---curated by LF Networking

2021

• 39. Machine Learning for finding groups of BSSs (Basic Service Set) suitable for Coordinated Spatial Reuse---curated by UPF
• 38. Cross Layer user experience optimization – Radio link performance prediction---curated by China Mobile

• 37. Combinatorial Optimization Challenge: Delivery route optimization---curated by ZTE

• 36. Federated Learning for Spatial Reuse in a multi-BSS (Basic Service Set) scenario---curated by UPF

• 35. Forecasting Model for Service Allocation Network Using Traffic Recognition---curated by SPbSUT

• 34. Graph Neural Networking Challenge 2021: Creating a Scalable Network Digital Twin---curated by BNN-UPC

• 33. Lightning-Fast Modulation Classification with Hardware-Efficient Neural Networks---curated by Xilinx

• 32. Location estimation using RSSI of wireless LAN---curated by RISING

• 31. ML5G-PHY-Localization: Multidevice localization with mmWave signals in a factory environment---curated by NC State University

• 30. ML5G-PHY-Reinforcement learning: scheduling and resource allocation---curated by UFPA

• 29. Network anomaly detection based on logs---curated by China Unicom

• 28. Network failure detection and root cause analysis in 5GC by NFV-based test environment---curated by KDDI

• 27. Build-a-thon(PoC) Network resource allocation for emergency management based on closed loop analysis---curated by ITU Focus Group on Autonomous Networks (FG-AN)

• 26. Radio Link Failure Prediction---curated by Turkcell

• 25. RF-Sensor Based Human Activity Recognition---curated by The University of Alabama

• 24. WALDO (Wireless Artificial intelligence Location DetectiOn): sensing using mmWave communications and ML.---curated by NIST

2020

• 23. Alarm and prevention for public health emergency based on telecom data---curated by China Unicom (Beijing Division)
• 22. Core Network KPI index anomaly detection---curated by China Unicom (Shanghai Division)

• 21. 5G+AI (Smart Transportation)---curated by JNU,IIT/Delhi

• 20. 5G+AI+AR (Zhejiang Division)--- curated by China Unicom
• 19. Analysis on route information failure in IP core networks by NFV-based test environment ---curated by KDDI
• 18. Compression of Deep Learning models---curated by ZTE

• 17.Demonstration of MLFO capabilities via reference implementations---curated by Letterkenny Institute of Technology, Co. Donegal

• 16. DNN Inference Optimization Challenges---curated by ADLIK, ZTE

• 15. Energy-Saving Prediction of Base Station Cells in Mobile Communication Network---curated by China Unicom
• 14. Fault Localization of Loop Network Devices based on MEC Platform ---curated by China Unicom

• 13. Improving the capacity of IEEE 802.11 WLANs through machine learning---curated by UPF

• 12. ML5G-PHY -Beam-Selection: Machine Learning Applied to the Physical Layer of Millimeter-Wave MIMO Sytems---curated by UFPA
• 11. ML5G-PHY- Channel Estimation @NCSU: Machine Learning Applied to the Physical Layer of Millimeter-Wave MIMO Systems at North Carolina State University--- curated by NC State University

• 10. Network State Estimation by Analyzing Raw Video Data--- curated by NEC

• 9. Network topology optimization --- curated by China Mobile
• 8. Out of Service(OOS) Alarm Prediction of 4/5G Network Base Station --- curated by China Mobile

• 7. Privacy Preserving AI/ML in 5G networks for healthcare applications--- curated by C-DOT, IIT/Delhi

• 6. Using Weather Info for Radio Link Failure Prediction Challenge--- curated by Turkcell

• 5. Shared Experience Using 5G+AI (3D Augmented + Virtual Reality)--- curated by Hike, IIT/Delhi

• 4. Traffic recognition and long-term traffic forecasting based on AI algorithms and metadata for 5G/IMT-2020 and beyond--- curated by SPbSUT

• 3. Graph Neural Networking Challenge--- curated by BNN, UPC

• 2. Improving experience and enhancing immersiveness of Video conferencing and collaboration--- curated by Dview

• 1. 5G+ML/AI (Dynamic Spectrum Access)--- curated by IITD

GeoAI Challenge

ITU GeoAI Challenge is a pioneering competition focused on leveraging artificial intelligence (AI) and machine learning (ML) to tackle real-world geospatial issues, particularly those aligned with the UN Sustainable Development Goals (SDGs). Beyond competing for prizes and recognition, participants gain valuable hands-on experience in AI/ML, directly contributing to solving critical SDG-related challenges. The GeoAI Challenge embodies a collaborative effort to innovate and apply geospatial intelligence towards a sustainable future. Below are the various problem statements hosted since 2022;

2023

• 8. GeoAI Challenge Location Mention Recognition from Social Media--- curated by QCRI, QU, Qen Labs Inc.

• 7. GeoAI Challenge Estimating Soil Parameters from Hyperspectral Images--- curated by ESA (European Space Agency)

• 6. GeoAI Challenge for Air Pollution Susceptibility Mapping--- curated by GEOlab at Polytechnic di Milano

• 5. GeoAI Challenge for Cropland Mapping--- curated by UNODC, FAO

• 4. GeoAI Challenge for Landslide Susceptibility Mapping--- curated by GEOlab at Polytechnic di Milano

2022

• 3. Cropland mapping with satellite imagery--- curated by FAO

• 2. Location Mention Recognition from Social Media Crisis-related Text--- curated by Qatar Computing Research Institute (QCRI, HBKU), and Qatar University (QU)

• 1. School mapping with big data--- curated by UNICEF

TinyML Challenge

tiny Machine Learning (ML) field involves applying machine learning to small, power-constrained devices and embedded systems. It encompasses hardware, algorithms, and software capable of on-device sensor data analytics at very low power, enabling various always-on applications. ITU and industry partners have hosted the tinyML Challenge since 2022, aiming to develop a cost-effective, low-power, and reliable smart weather station without mechanical moving parts. The Challenge also explores scalable and high-performance tinyML solutions uses cases relevant to SDGs. Problem statements are listed below:

2024

• 5. Next-Gen tinyML Smart Weather Station Challenge--- curated by CSEM

2023

• 4. Next-Gen tinyML Smart Weather Station--- curated by CSEM, tinyML Foundation

• 3. Scalable and High-Performance TinyML Solutions for Plant Disease Detection--- curated by ITU

• 2. Scalable and High-Performance TinyML Solutions for Wildlife Monitoring--- curated by ITU

2022

• 1. Smart Weather Station Challenge---curated by TinyML Foundation

AI for Fusion Energy Challenge

The Challenge explored the potential of ML in enabling predictive modelling for fusion energy systems. Through this Challenge, participants developed a cross-machine disruption prediction model using ML, with strong generalization capabilities. One problem statement was offered in 2023 below:

2023

• 1. Multi-Machine Disruption Prediction Challenge for Fusion Energy---curated by ITU, IAEA, PSFC, HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY