| 摘要: |
| 针对大规模复杂网络及其特殊场景下无线网络规划存在需求精准定位难、投资效益评估不全面、人工工作量大等痛点，提出了数据与知识联合驱动的网络规划技术体系，基于多模态数据构建规划数据底座，利用高精度孪生仿真提前预知规划方案效果，通过无线规划智能体重塑规划工作流程，运用系列化数智能力筑牢无线网络自智规划底座能力，实现了网络规划从传统人工模式向自动化与智能化模式的转型升级。此外，还进行了无线网络自智规划的示范应用，试验结果表明在典型乡农场景通过工参调优区域覆盖率提升20.27%，典型城市场景通过自智规划平均网络性能提升167 dB且达到规划目标，验证了所提网络规划技术体系的科学性和精准性，为无线网络的高质量可持续规划建设提供了可靠的技术路径与实践范例。 |
| 关键词: 通信网络规划 数据与知识 孪生仿真 智能体 自智网络 |
| DOI：10.20079/j.issn.1001-893x.251205002 |
|
| 基金项目: |
|
| Technical System of Network Planning Driven by Integration of Data and Knowledge |
| ZHANG Shu,DIAO Lei,CHE Chen,LI Bin,XIANG Zhongqiu,TANG Jun,YAN Yaofei,MA Yonggang |
| (1.China Mobile Group Yunnan Co.，Ltd.，Kunming 650228，China;2.China Mobile Group Design Institute Co.，Ltd.，Beijing 100080，China) |
| Abstract: |
| Expanding the coverage breadth and depth of mobile networks in frontier regions，remote mountainous areas and other scenarios has become a key priority. For large-scale complex networks and their special application scenarios，wireless network planning is plagued by such pain points as difficulties in accurate demand positioning，incomplete investment benefit evaluation，and heavy manual workload. The authors propose a data and knowledge co-driven technical system for network planning.A planning data foundation based on multi-modal data is constructed，the performance of planning schemes is predicted in advance through high-precision digital twin simulation，the planning workflow is reshaped with intelligent wireless planning agents，and the foundational capabilities of autonomous and intelligent wireless network planning are consolidated by leveraging a series of data-intelligent capabilities，thus realizing the transformation and upgrading of network planning from the traditional manual mode to an automatic and intelligent mode. Furthermore，a demonstration application of intelligent autonomous planning is carried out for wireless networks.The test results show that in typical rural scenarios，the regional coverage rate is improved by 20.27% through engineering parameter optimization；in typical urban scenarios，the average network performance is enhanced by 1.67 dB via intelligent autonomous planning and meets the planning objectives.This verifies the scientificity and accuracy of the proposed technical system，and provides a reliable technical approach and practical paradigm for the high-quality and sustainable planning and construction of wireless networks. |
| Key words: communication network planning data and knowledge digital twin simulation agent autonomous network |
《电讯技术》编辑部 