Session

Computer Science and Communication Engineering

Description

As 5G era is approaching fast and pre-commercial 5G tests and trials are happening everywhere around the world, one of the key challenges for carriers and 5G providers is to maintain and operate the network complexity required to meet diverse services and personalized user experience requirements. This maintenance and operation have to be smarter and more agile in 5G than it was in previous generations. AI and ML can be leveraged in this case to ease 5G complexity and at the same time enhance the intelligent connectivity between diverse devices and diverse tiny end points, e.g. IoT sensors.

Machine learning and AI algorithms can be used to digest and analyse cross-domain data that would be required in 5G in a much more efficient way enabling quick decision and as such easing the network complexity and reducing the maintenance cost. The cross-domain data includes geographic information, engineering parameters and other data to be used by AI and ML to better forecast the peak traffic, optimize the network for capacity expansion and enable more intelligent coverage through dynamic interference measurements.

This paper provides an overview of 5G complexity due to its heterogeneous nature and the key role of AI and ML to ease this complexity and enhance the intelligent connectivity between diverse devices with different requirements. The focus of this paper will be on the key aspects of AI and ML application in 5G and the key benefits from this application. Finally, this paper will analyse the overall performance of 5G in terms of coverage and latency compared with traditionally operated networks.

Keywords:

5G, AI, ML, IoT, Sensors, Network Slicing, Virtualisation, MTC.

Session Chair

Xhafer Krasniqi

Session Co-Chair

Driart Elshani

Proceedings Editor

Edmond Hajrizi

ISBN

978-9951-550-19-2

First Page

45

Last Page

51

Location

Pristina, Kosovo

Start Date

26-10-2019 3:30 PM

End Date

26-10-2019 5:00 PM

DOI

10.33107/ubt-ic.2019.278

Download

Share

COinS
 
Oct 26th, 3:30 PM Oct 26th, 5:00 PM

AI leverage in easing the 5G complexity and enhancing 5G intelligent connectivity

Pristina, Kosovo

As 5G era is approaching fast and pre-commercial 5G tests and trials are happening everywhere around the world, one of the key challenges for carriers and 5G providers is to maintain and operate the network complexity required to meet diverse services and personalized user experience requirements. This maintenance and operation have to be smarter and more agile in 5G than it was in previous generations. AI and ML can be leveraged in this case to ease 5G complexity and at the same time enhance the intelligent connectivity between diverse devices and diverse tiny end points, e.g. IoT sensors.

Machine learning and AI algorithms can be used to digest and analyse cross-domain data that would be required in 5G in a much more efficient way enabling quick decision and as such easing the network complexity and reducing the maintenance cost. The cross-domain data includes geographic information, engineering parameters and other data to be used by AI and ML to better forecast the peak traffic, optimize the network for capacity expansion and enable more intelligent coverage through dynamic interference measurements.

This paper provides an overview of 5G complexity due to its heterogeneous nature and the key role of AI and ML to ease this complexity and enhance the intelligent connectivity between diverse devices with different requirements. The focus of this paper will be on the key aspects of AI and ML application in 5G and the key benefits from this application. Finally, this paper will analyse the overall performance of 5G in terms of coverage and latency compared with traditionally operated networks.