基于深度学习的5G覆盖动态优化的研究与应用

Research and Application of 5G Coverage Dynamic Optimization Based on Deep Learning

针对传统覆盖优化评估难、效率低、准确性差的问题,提出了5G覆盖动态优化方法。首先,针对覆盖评估,创新构建了5G MR切片分析诊断体系。其次,在5G用户三维空间定位方法上探索研究了基于DOA的用户分布估算。进一步地,提出了基于MR和DOA数据的Massive MIMO权值智能寻优方案,深度挖掘5G覆盖潜力。目前整套方法已融入日常优化工作,在网络覆盖的智能诊断和自动调优中发挥着核心作用,为构建客户网络感知满意的优选运营商起到了重要推进作用。

Aiming at the problems of difficult evaluation,low efficiency and poor accuracy of traditional coverage optimization,the 5G coverage dynamic optimization method is proposed. Firstly,5G MR slice analysis and diagnosis system is innovatively constructed for 5G coverage evaluation. Secondly,the user distribution estimation based on DOA is explored and studied on the three-dimensional spatial positioning method of 5G users. Further,the massive MIMO weight intelligent optimization scheme based on MR and DOA data is proposed to deeply mine the 5G coverage potential. At present,the whole system has been put into daily optimization,which plays a core role in intelligent diagnosis and automatic optimization of network coverage and plays an important role in promoting the construction of preferred operators with customer network perceived satisfaction.