SKT Applies Network Quality Management 'A-STAR' to Base Stations Nationwide

SK Telecom announced on the 28th that it has developed the wireless network quality management AI solution A-STAR (Access-Infra Service for Targeting & Action Recommendation) and applied it to its nationwide base stations.

A-STAR continuously monitors the status of hundreds of thousands of base stations nationwide, identifies base stations expected to affect customer perceived quality, notifies field operation personnel, and simultaneously recommends the causes of issues and improvement methods.

The solution consists of functions such as 'Wireless Quality Monitoring,' which analyzes quality data of nationwide base station equipment every hour and reports anomalies to personnel; 'Cause Analysis of Quality Degradation,' which analyzes about 250 types of quality data per device to identify and provide the core causes of degradation; and 'Improvement Recommendation,' which suggests appropriate improvement methods for quality degradation data.

SKT reported that compared to previous methods, the use of A-STAR for field quality improvement activities last year reduced the time required for quality analysis by 80.7% and improved proactive base station improvements in the field by approximately 46.7%.

SKT began developing A-STAR in 2020 and secured the practical usability of the AI model through two years of development, field activity verification, and gradual functional enhancements.

In cases where recommendation results are inappropriate, the final actions taken by field quality improvement personnel are fed back for retraining, enabling continuous improvement of AI performance. Over time, more accurate information can be provided.

SKT plans to continuously advance A-STAR by shortening the hourly quality analysis cycle and expanding optimal route recommendation functions to enhance the effectiveness of field personnel’s improvement activities.

Park Myung-soon, head of SKT Infrastructure DT, said, "SKT has upgraded network operation intelligence by utilizing AI technology in customer quality management activities," adding, "We will continue to evolve as an AI infrastructure by sustaining changes in the field."