基于GSM手机测量统计无线网络评估方法.docVIP

基于GSM手机测量统计的无线网络评估方法 摘要 在GSM无线网络优化中，无线电波的信号强度、覆盖情况是整体无线网络的性能的关键。对于广大的手机用户来说，能够在何时何地都有良好的信号覆盖，保持正常的通话需求，是我们作为移动通信运营商的使命。 要掌握空间中每一个地方的信号情况，我们必须要通过测试或者场强预测的手段来了解。 测试通常是使用测试设备到现场获取无线电波的能量情况。这种方法有一定的局限性，路测可以较为真实的反映无线信号覆盖情况，但只能对道路进行测试和评估。对于相当多的没有道路的住宅等区域，大量的测试并不现实。 而场强预测则是通过套用一些国内外通用的无线电波统计传播模型进行计算，从而得出结果。国内外常用的统计无线电波传播模型为著名的Okumura-Hata即奥村模型，和Cost231-Hata模型。 我们在工程中进行场强预测通常采用规划工具（例如爱立信TCP）。利用规划工具虽然可以对网络中所有点进行覆盖评估，但这种评估是一种静态的评估，网络或实际无线环境发生变化后（规划所用地图难以更新）就会使得评估跟不上变化，从而产生较大误差。有传播模型滞后，传播模型单一等缺点。 本文试图利用爱立信OSS系统的网优统计数据（MRR，NCS）以及小区配置参数，研究对无线网络的传播模型进行动态配置和修正评估的方法。在计算的过程中通过采用样本修正、MMSE线性拟合等方法对传播模型进行修正，修正后的传播模型可实现计算区域内各处的信号场强。 此方法已经应用到广州无线网络优化科技项目中，并且在现网中得到了验证。 关键词：无线覆盖评估 MRR NCS 路径损耗方程 MMSE Abstract In the field of GSM radio network optimization, the radio signal strength and radio coverage is the key index. It’s the very duty for the mobile communication operator to keep every mobile in the situation of good signal and quality. We must make clear of the radio coverage of the whole place, by means of field testing or coverage prediction. Usually we get the coverage situation by drive test. But there is one limitation of the drive test, that is you can never get the coverage of indoor situation, where there is no way to drive. Coverage prediction is usually done in the method of the famous radio propagation models, which are the Okumura-Hata model and Cost231-Hata model. We have the planning tool(like Ericsson TCP) to do the coverage prediction. But it’s kind of a still prediction with some weakness as lacking of fitting the propagation parameters. In this paper , we were trying to fine tune the radio propagation parameters to achieve a better way to predict the radio coverage. Sample fitting and MMSE curve fitting are introduced to do this fine tuning. The tuned propagation can easily calculate the coverage situation of every GSM site. The algorithms were testified and applied in the project of GuangZhou Mobile GSM radio network optimization. Key Words：Coverage Prediction, MR