Fifth Generation Wireless Channels Outlier Detection and Clustering

The fifth generation (5G) wireless communications system offers faster data rates, lower latency, and higher number of interconnecting devices. Various 5G channel models were developed to study its stochastic characteristics prior to its implementation. These channel models generate multipath components that are grouped into clusters when they have similar properties in delay and angles. The multipaths and multipath clusters are used as datasets in multipath clustering which is used to examine the propagation properties of the 5G system. However, datasets are prone to outliers. They tend to affect clustering accuracy. Hence, this study clusters the datasets generated by the channel models, remove the outliers, and cluster again the datasets free of outliers. Outlier detection shows 5G channel model datasets contain noise and outlier removal improves the modelling characteristics shown by improved clustering accuracy. Results show that most of the outliers are detected in the 2*SD theshold. The removal of the outliers increased the clustering accuracy. This shows that outlier detection and removal also work well with channel model datasets and can be used in analyzing the propagation characteristics of 5G.

第五代（5G）无线通信系统提供了更快的数据传输速率、更低的延迟以及更高的设备连接数量。为了在实施前研究其随机特性，开发了多种5G信道模型。这些信道模型生成的多径成分在延迟和角度相似时，会被聚集成簇。这些多径及其簇被用作多径聚类数据集，用以考察5G系统的传播特性。然而，数据集易受异常值的影响，它们往往会影响聚类的准确性。因此，本研究对信道模型生成的数据集进行聚类，去除异常值，并对无异常值的数据集再次进行聚类。异常值检测表明，5G信道模型数据集中存在噪声，去除异常值提高了由改进的聚类准确性所展现的建模特性。结果显示，大部分异常值均在2倍标准差阈值内被检测到。去除异常值提高了聚类准确性，这表明异常值检测与去除在信道模型数据集中同样适用，并且可用于分析5G的传播特性。