Background:
Massive MIMO, as one of the key features of 5G, can change the horizontal and vertical coverage of 5G cells by adjusting the antenna weights, which can effectively increase the coverage distance and reduce the interference between users. Massvie MIMO uses a large-scale antenna array, which greatly increases the space for antenna parameter adjustment. There are thousands of beam combinations for different SSB beam parameter configurations, and the traditional manual experience method cannot be applied in the 5G field.
At present, most of the SSB beam optimization algorithms do not dynamically adjust the beam configuration according to the user distribution, and because there is no interactive simulation platform, the optimization objectives of the existing algorithms are mostly SS-RSRP and SS-SINR, the network side indicators are not introduced as the optimization objective. SSB beam optimization is expected to improve the performance of the network by improving, solving coverage or quality issues. Therefore, based on this background, a multi-objective joint antenna optimization task is proposed, and the performance index of the network side is added as the optimization goal.
Multi-objective joint antenna optimization needs to design and train an optimization algorithm and model according to the provided interactive wireless network simulation environment and user distribution data, which can dynamically output the optimal beam configuration according to the user distribution. The optimization target not only includes SS-RSRP, SS-SINR, but also includes the number of access users based on the network output under the current beam configuration, downlink traffic, downlink rate and other network side performance metrics.
Tasks:
1.Ai-based Multi-Objective Antenna Optimization
Expected Outcomes:
1.The network coverage capacity index optimization model applicable to multiple scenarios：For 5G MassiveMIMO site networking environment, combined with the change of user location distribution, the network coverage and capacity indicators such as regional coverage, rate and throughput were jointly optimized.The optimization effect and efficiency can reach the industry-leading level.
Deadline:
By March 30, 2025
Potential Respondents:
Beijing University of Posts and Telecommunications,
Zhejiang University of Technology, Hangzhou Dianzi University
About GTI
GTI is an international cooperation platform initiated and established by China Mobile, Softbank, Vodafone and other operators in 2011. It currently has 146 operator members and 258 industry partners. In 2023, the new stage of GTI 3.0 was officially launched, and it is committed to continuously deepening the global cooperation of 5G-A+AI and achieving win-win commercial success. For more information, please visit http://gtigroup.org/
About the GTI 5G-A x AI Open Development Program
In Feb 2024 at MWC Barcelona, GTI launched the 5G-A×AI Development Program to promote the integrated innovation of 5G and AI in technology, business, ecology, and commerce, and two-way empowerment. Therefore, 5G is smarter and AI is more ubiquitous, which will support the goals of GTI 3.0. First, Build Open Labs to provide basic environment, equipment facilities, industry application scenarios and other resources for 5G-AxAI integration innovation, and carry out the R&D, testing and demonstration of new technologies and solutions. Second, Build an Open Collaborative Innovation Community, with an online platform for “Communication and Sharing” and “Supply and Demand Matching”, and jointly carry out cutting-edge exploration, technical research, testing and iterative optimization. Third, Explore Innovative 5G-AxAI Integration Use Cases, and condense themreplicable business model templates, so as to provide references for value creation and monetization.
