Instructors
Dr. Melike Erol-Kantarci, Melike.erolkantarci@uottawa.ca
University of Ottawa, Canada
Melike Erol-Kantarci is Canada Research Chair in AI-enabled Next-Generation Wireless Networks and Associate Professor at the School of Electrical Engineering and Computer Science at the University of Ottawa. She is the founding director of the Networked Systems and Communications Research (NETCORE) laboratory. She is a Faculty Affiliate at the Vector Institute. She has over 150 peer-reviewed publications which have been cited over 5500 times and she has an h-index of 38. Recently, she received the 2020 Distinguished Service Award of the IEEE ComSoc Technical Committee on Green Communications and Computing and she was selected among the N2Women Stars in Computer Networking and Communications in 2019. She has received several best paper awards including the IEEE Communication Society Best Tutorial Paper Award in 2017. Dr. Erol-Kantarci is the co-editor of three books on smart grids, smart cities and intelligent transportation. She has delivered 50+ keynotes, plenary talks, tutorials and panels around the globe. She isthe Specialty Chief Editor for the Smart Grid Communications Section of the journal “Frontiers in Communications and Networking”. She is on the editorial board of several IEEE journals. She has acted as the general chair and technical program chair for many international conferences and workshops. She is the vice-chair for IEEE ComSoc emerging technologies initiative on Machine Learning for Communications. She is a steering committee member for the IEEE Sustainable ICT Initiative. Her main research interests are AI-enabled wireless networks, 5G and 6G wireless communications, Open Radio Access Networks (ORAN), smart grid, electric vehicles, Internet of things and wireless sensor networks. She is an IEEE ComSoc Distinguished Lecturer, IEEE Senior memberandACM Senior Member.
Dr. Meryem Simsek, msimsek@vmware.com
VMware Inc., USA
Dr. Meryem Simsek is a chief scientist of intelligent networking at VMware. Prior to this, she was a senior research scientist at Intel Labs, where she worked on machine learning solutions for 5G and beyond and was involved in various NSF programs. In her roles as senior research scientist at ICSI Berkeley and as research group leader at TU Dresden, she has been actively working on 5G and beyond, self-organizing networks, and on the Tactile Internet, such that she is a well-known scientist in both academia and industry. She has been working on Machine Learning for wireless systems since a decade and is the recipient of the IEEE Communications Society Fred W. Ellersick Prize in 2015 and the Rising Star in Computer Networking and Communications by N2Women in 2019. She has over 80 peer-reviewed publications, has co-edited two books, and has delivered 30+ invited talks and keynotes. She has initiated and is the Chair of the IEEE Tactile Internet Technical Committee and serves as the Vice Chair for the IEEE P1918.1 Standardization Working Group, which she has co-initiated. Additionally, she is a member of the IEE ComSoc Strategic Planning Committee, the IEEE ComSoc Emerging Technologies Committee, the vice-chair of the IEEE FNI Systems Optimization Working Group, and the coordinator of industry and student activities of the IEEE Women in Communications Engineering Committee. Her main research interests include future wireless networks, intelligent networking, machine learning, and self-organizing networks. She is a senior member of IEEE.
Motivation
Future wireless networks, i.e. 5G and the upcoming 6G, are expected to simultaneously accommodate diverse use cases. In addition, resource efficiency, reliability, and robustness are becoming more stringent for 5G and beyond networks. To meet this, 6G must incorporate a paradigm shift in network and radio resource optimization, in which efficient and intelligent resource management techniques have to be employed. In addition to all those new types of services and demands, wireless networks are at the cusp of a new paradigm with open virtualized architectures where softwarization of networks is helping to disaggregate network functions in the wireless domain and allowing for ultimate autonomy capabilities. Artificial intelligence (AI), or more specifically machine learning (ML) algorithms stand as promising tools to intelligently manage the networks such that network efficiency, reliability and robustness goals are achieved, quality of service demands are satisfied, network and computational resources are used most efficiently, and performance targets are achieved in a self-optimized manner. The opportunities that arise from learning the environment parameters under varying conditions, positions AI-enabled 5G and 6G superior to preceding generations of wireless networks. In addition to using AI for networks, the distributed nature of networks provides a natural environment for enhanced machine learning opportunities. This tutorial will begin with an introduction to 5G and beyond networks together with open radio access network (RAN) features and some fundamentals on ML. After summarizing the state-of-art in ML algorithms and their applications to (open) RAN, it will continue with a full-fledged treatment of clustering algorithms, reinforcement learning, deep learning,and federated learning techniques. Finally, challenges and open issues will be discussed both in terms of AI algorithms and their applicability to various functions of future wireless networks. These discussions will be put into perspective considering the recent 5G NR Releases and the plans for 6G.
Structure and content
Over the past decade, the huge growth in data across many different fields resulted in big data challenge which amplified the need for intelligent data analysis schemes. Various machine learning methods emerged, such as deep learning, and they have been used along with traditional machine learning methods to cope with the big data problem. Recently they have been adopted in wireless networks. There is a growing literature in AI-enabled wireless networks, but yet a growing gap of knowledge between fundamentals of machine learning and studies that blindly adopt well-known algorithms to wireless network problems. This tutorial aims to fill this gap by giving a comprehensive treatment of the literature on both sides of the topic, which are advances in machine learning and advances in wirelessnetworks.
The tutorial will be in lecture style and it is structured as follows:
- Introduction to 5G (Release 16+) and ORAN (15 mins)
- 5G NR
- ORAN and the RAN Intelligence Controller
- Fundamentals of ML (15 mins)
- Unsupervised learning
- Supervised learning
- Reinforcement learning
- State-of-the-art in studies using Machine Learning (ML) for wireless (30 mins)
- Over 50 works will be summarized from the past 10 years
- Studies both from LTE and 5G domain will be introduced
- ML applications to wireless RAN problems (30 mins)
- Unsupervised learning (k-means, DBSCAN as in used for traffic classification)
- Reinforcement learning (Q-learning, actor-critic learning as in used for interference mitigation, spectrum allocation)
- Convolutional networks and deep learning (LSTM, Deep Q-learning as in used for resource allocation)
- Edge Intelligence and applications (45 mins)
- Federated learning
- Staleness Control
- New Trends and how AI can have an impact on 6G (15 mins)
- Release 16 and Release 17
- AI-enabled network vision towards 6G
- Open Issues and Future Directions (15 mins)
- Q&A (15 mins and more)