SK텔레콤은 국내 통신사 최초로 유선망 운용에 필요한 모든 제어·점검 작업의 자동화가 가능한 차세대 유선망 제어 플랫폼 ‘AI 오케스트레이터(AI Orchestrator)’를 자체 개발하고 5G/LTE 유선망 전체에 적용했다고 15일 밝혔다.
Infrastructure-as-a-Code (IaC) solution applied to 5G/LTE wired networks
AI model development and analysis… Network operation efficiency and stability↑
SKT unveiled a code-type infrastructure solution to automate wired network operation management
SK Telecom announced on the 15th that it has developed its own next-generation wired network control platform, 'AI Orchestrator', which is the first domestic telecommunications company to automate all control and inspection tasks required for wired network operation, and has applied it to its entire 5G/LTE wired network.
AI Orchestrator is a platform that automates software-defined networks (SDNs) that operate networks through infrastructure as code (IaC). When network operators input scripts in a programming language they are familiar with, they are automatically translated into commands for each device, enabling integrated inspection and control of tens of thousands of devices nationwide.
In order to control the wired network consisting of Internet Protocol (IP)-based communication equipment from individual base stations to the backbone network, it took a long time to learn all the different commands for each equipment and model. To solve this problem, we have been promoting data model standardization and developing and using SDN technology for equipment control, but there have been several limitations, such as high maintenance costs and differences in standardization application by equipment manufacturer.
SKT said, “The automated system of the AI orchestrator not only overcame existing limitations by enabling integrated control of the entire equipment, but also significantly improved work speed, such as completing the work in just one day. Errors that could occur during manual management were also reduced.”
In addition, the AI orchestrator facilitates version management and change tracking of wired networks, enabling quick recovery from problems, thereby enhancing the stability of network operation. It automatically detects codes that execute dangerous commands, and when work is required with such codes, it enables execution after approval from the approval authority, thereby enhancing the security of network operation.
Even when an error occurs in the control and inspection application that is the basis of the operational function, unlike the past when the main software had to be modified and the build and deployment process had to be completed, AI Orchestrator can quickly complete the deployment by simply modifying the template code.
SKT announced that it has embedded AI-based analysis and model functions here. This allows for precise analysis and prediction of network operations. SKT developed a data analysis system that collects data continuously generated from telecommunications equipment installed nationwide, and applied a deep learning framework to the platform so that network operators can easily utilize AI.
Through this environment, network operators are expected to be able to develop AI models based on stored data, analyze the data to determine what actions and inspections are needed in the future, and increase the accuracy of network anomaly detection and traffic prediction.
Meanwhile, SKT also introduced AI Orchestrator and had it adopted as a controller task at the IOWN, a global next-generation communications standards forum held in Canada for three days from April 23. SKT will document the AI Orchestrator and continue to evolve it to strengthen its network technology leadership.
SKT Ryu Tak-ki, head of infrastructure technology at SKT, said, “This is very meaningful in that we are the first domestic telecommunications company to introduce NetDevOps for wired network operations,” and added, “As an AI company, we will take the lead in developing solutions to apply AI to our core infrastructure area, as well as establishing a development culture.”