Page 357 - AI for Good Innovate for Impact
P. 357
AI for Good Innovate for Impact
Use Case 14: AI-Powered Network O&M, Ensuring Uninterrupted
Network During Power Rationing 4.3 - 5G
Organization: MTN Group
Country: Southern Africa
Contact Person(s):
Francis Yang, yangmingxi1@ huawei .com
Nkosinathi, Nkosinathi.Nzima@ mtn .com
1 Use Case Summary Table
Item Details
Category 5G
Due to network congestion caused by power rationing, the packet loss
Problem Addressed rate increases on the bearer network, and the invisible traffic suppres-
sion leads to inaccurate network construction, poor user experience
The main elements of the solution include AI-enabled Network Elements
Key Aspects of Solu- (NE), a network digital map [3][4], and a network optimization agent.
tion An application based on the telecom foundation model [2] enables the
automation of network analysis and optimization
SRv6 (Segment Routing over IPv6), Network Digital Map, NL2SQL (Natu-
Technology Keywords
ral Language to SQL), Network Awareness
Data Availability Not Available
Metadata (Type of Text, including the network object type, performance indicator, and
Data) indicator description/explanation, etc.
Model Training and Telecom foundation model based on transformer fine-tuning, 38 billion
Fine-Tuning parameters, natural language corpus
Testbeds or Pilot Internal Testbed is used inside the organisation
Deployments
Code repositories Not Available
321
