基于大数据的移动网络测评系统设计和优化

Design and Optimization of Big Data-based Mobile Network Evaluation System

提出了基于Spark、HBase和JavaWeb的移动网络测评系统设计,把栅格数据进行地域分组聚合和比邻分区存储,提供适合分布式并行处理的数据架构,设计自主研发的集群,采用完全分布式架构完成从数据检索到测评切片子图生成的全过程,实现了全高清分辨率下像素级栅格化测评结果的高效可视化。

Based on Spark, HBase and JavaWeb, this paper proposes a design of mobile network evaluation system. The grid data is grouped by region and stored by partition, a data structure is provided to be suitable for distributed parallel processing, and an independently developed cluster is designed. The entire procedure from data query to evaluation slice sub-image generation is completed by a fully distributed architecture, and efficient visualization is realized with pixel-level grid evaluation under full-HD screen resolution.