BUPTCMCC-6G-DataAI+: A generative channel dataset for 6G AI air interface research

Li Yu,Jianhua Zhang,Mingjun Fu,Qixing Wang
2024-09-30
Abstract:In September 2024, Beijing University of Posts and Telecommunications and China Mobile Communications Group jointly releases a channel dataset for the sixth generation (6G) mobile communications, named BUPTCMCC-6G-DataAI+. BUPTCMCC-6G-DataAI+ is the update version of BUPTCMCC-6G-DataAI, which is already published in June 2023, aiming at extending 6G new technologies, frequency bands, and applications. BUPTCMCC-6G-DataAI+ provides deterministic data covering new mid-bands, millimeter wave (mmWave), and terahertz (THz), supports the features of XL-MIMO near-field, high mobility and provides multiple 6G scenarios such as reconfigurable intelligent surface (RIS) and industrial Internet. Configured with customized features according to different user needs, BUPTCMCC-6G-DataAI+ can adaptively generate scalable large-scale or small-scale parameters, providing data support for 6G research and development, and standardization.
Signal Processing
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to provide a more comprehensive, flexible and customized channel data set for the research of the sixth - generation (6G) mobile communication system. Specifically, the BUPTCMCC - 6G - DataAI+ data set aims to solve the following problems: 1. **Support for new frequencies and technologies**: - The new mid - band, millimeter - wave (mmWave) and terahertz (THz) bands bring more abundant spectrum resources and larger bandwidth, thus increasing the transmission rate. The existing data sets do not fully cover these new bands and new technology features. 2. **Integration of artificial intelligence and communication**: - The Radiocommunication Sector of the International Telecommunication Union (ITU - R) lists the integration of AI and communication as a typical application scenario for 6G. In order to design and optimize 6G communication algorithms (such as beam management in the physical layer, power allocation in the resource layer, and network planning and optimization in the network function layer), more specific and sufficient training data are required. 3. **Generate customized data**: - Most of the existing data sets are generated based on statistical models, channel measurements or deterministic scenario simulations, but these data sets lack some important propagation mechanisms (such as diffraction), which are crucial for some algorithm designs (such as long - distance blockage prediction). In addition, the existing data sets fail to consider the new features of 6G, such as XL - MIMO with near - field spherical wave characteristics, reconfigurable intelligent surface (RIS) and sensing scenarios. 4. **Provide support for diverse scenarios**: - BUPTCMCC - 6G - DataAI+ provides support for multiple scenarios, including outdoor street scenarios, indoor office scenarios, indoor RIS scenarios, industrial Internet scenarios and outdoor urban scenarios, etc., to meet different research needs. 5. **User - customized configuration**: - This data set allows users to flexibly configure parameters, such as base stations, user areas, antennas, scenarios and frequencies, and can choose whether to enable functions such as mobile interpolation, RIS and dynamic scenarios, etc., so as to generate specific data on a large or small scale. By solving the above problems, the BUPTCMCC - 6G - DataAI+ data set can provide more diverse and customized data support for the research and development of 6G wireless communication systems, promote the in - depth integration of AI and communication, and drive the innovation and development of 6G technology.