5G Smart Networks Part 1: Network Slicing

In this video, Dr Natalie Whitehead explains what 5G network slicing is, and we find out how researchers at the University of Exeter are using AI and Machine Learning techniques to improve how resources can be shared out in 5G and B5G (beyond 5G) networks.

To find out more, check out the website we’ve created for this research team: and see the second video of the series here: https://youtu.be/QIe8JZctZM0

This educational video is funded through the team’s research grant as part of our Public Engagement Programmes (https://exetersciencecentre.org/help/researchers-info/#impact) to help share their research with the public.

In this video we include a clip from Huawei / Rainforest Connection – note that 2G and 3G is currently used for their Rainforest Guardian systems, but 5G will likely be an enabler for future versions of their technology. To learn more about their work, you can watch the original video and read more here: https://www.huawei.com/minisite/tech4all/en/rainforest.html

— Chapters —
00:00 Introduction
00:40 What is 5G & how is it different to 4G?
01:18 User demands in a 5G network + network slicing
02:17 Network slicing: current challenges
03:19 Meet the researchers
03:43 Dr Yulei Wu – why do you use AI for network slicing, why not use a better algorithm to allocate resources?
05:53 Dr Haozhe Wang – what is the difference between machine learning and deep learning?
07:03 How does deep learning work?
08:42 What is reinforcement learning?
09:07 What is deep reinforcement learning?
09:46 Deep reinforcement learning applied to network slicing
11:24 Dr Yulei Wu – what other machine learning models are you using & what other problems are you trying to address with this technology?
12:10 Graph Neural Networks (GNN) for non-Euclidian data (such as network data)
12:43 Brief insight into Digital Twin technology using GNNs

Look out for our next video on Digital Twins – coming soon!