KT said on Oct. 28 it has received “Autonomous Network Level 4” certification from global telecoms industry body TM Forum in the field of optimising energy efficiency for 5G radio networks.

TM Forum is a consultative body that develops standards and digital transformation methodologies for the telecoms industry. The autonomous network certification assesses levels of autonomy based on how roles are divided between people and systems in network operations.

KT said Level 4 means AI understands network operating goals and autonomously performs everything from situation analysis to decision-making, execution and issue response. It said its in-house AI autonomous network operations platform, ES-O (Energy Saving Operation & Orchestration), played a key role in the certification.

ES-O analyses goals entered by an operator in everyday language and converts them into network operating policies. It analyses service coverage by base station and their mutual impacts to find combinations of stations that can improve energy efficiency while maintaining communications quality.

Before executing a policy, it analyses the latest traffic to check for possible quality degradation. If risk is detected, it automatically adjusts the scope and method of application. After execution, it continues to monitor network conditions and takes necessary measures.

It also learns information on adjacent base stations and relationships between base stations to predict traffic changes by time period. It maintains performance by comparing forecasts with actual traffic and retraining on the latest data if accuracy declines.

Sun-il Jeong (정선일), executive director and head of KT’s Network Operations Innovation Division, said, “We will continue to advance AI-based autonomous network technology while putting top priority on communications quality and network stability as experienced by customers.”