KT, in collaboration with Qualcomm and Rohde & Schwarz, demonstrates 6G AI technology.

• AI enables more accurate Channel State Information (CSI), improving downlink speed by 50%
• Collaboration with Qualcomm and Rohde & Schwarz proves AI-RAN performance in a commercial network environment
• KT: “We will strengthen global cooperation to evolve toward intelligent networks in the 6G era.”

KT (CEO Young-Shub Kim) announced on the 11th that it has successfully demonstrated an AI-based next-generation wireless transceiver technology in collaboration with global wireless technology leader Qualcomm Technologies Inc. (hereafter Qualcomm) and German communications equipment company Rohde & Schwarz.

The AI-based next-generation wireless transceiver technology is an AI-driven CSI (Channel State Information) compression technology currently being advanced under the international standardization body 3GPP Release 20.

By learning the complex radio environment between base stations and user devices, AI significantly improves the accuracy of CSI transmitted from devices to base stations. This ultimately enables the base station to support enhanced downlink speeds.

For this demonstration, KT and Qualcomm built a test environment integrating Qualcomm’s next-generation Snapdragon mobile platform with built-in AI capabilities and Rohde & Schwarz’s AI base station system (CMX500 OBT).

KT further enhanced the technical completeness of the solution by applying an AI model optimized for real-world communication environments, leveraging its operational experience in 5G networks and base station configuration data. The demonstration confirmed improved CSI accuracy and approximately a 50% enhancement in downlink performance.

This demonstration is significant as it shows that AI can directly improve wireless performance itself. In particular, it proves the effectiveness of 3GPP standards-based AI-RAN (intelligent base station) technology in a commercial network environment.

KT has been continuously developing AI technologies for network operations. In addition to this next-generation wireless transceiver technology, the company has developed and deployed “AI TEMS,” an AI-based optimal temperature management solution for central offices, and “ESKIMO,” a power-saving solution for base stations and servers.

Building on this demonstration, KT plans to further strengthen cooperation with global big tech companies to lead AI-based wireless technology development and 6G standardization. The company is also preparing for the evolution toward AI-based autonomous network operations.

Sangpyo Kim, President of Qualcomm Korea, stated, “6G will become an innovative platform for expanded intelligent edge AI, where AI controls network resources. Our collaboration with KT is an important step toward realizing this vision and is expected to accelerate the advancement of next-generation communication technologies.”

Andreas Pauly, CTO of Rohde & Schwarz, said, “We are very pleased to collaborate with KT in Korea, a leader in the global telecommunications market, to shape the future of 6G AI networks. Confirming the dramatic performance improvements achieved through AI-RAN and device solutions in this 6G AI demonstration marks an important milestone for future communication technologies.”

Jong-sik Lee, Executive Vice President and Head of KT’s Future Network Research Institute, stated, “6G is not merely about higher speeds, but about evolving into an intelligent network where AI and wireless communications are combined. Through this collaboration with global big tech companies Qualcomm and Rohde & Schwarz, KT will continue to secure wireless AI technologies to enhance the customer experience.”

* CSI (Channel State Information): In wireless communications, CSI refers to information that quantitatively describes the current quality, gain, and interference conditions of a wireless channel as measured by a base station or user device. In MIMO (Multiple-Input Multiple-Output) systems, CSI is used for channel estimation, beamforming, transmission mode selection, and scheduling optimization. Typically, the user device measures the channel and feeds the information back to the base station.