关键词: 城市计算, 物联网, 大数据, 异常过滤, 移动边缘计算

Abstract: In modern cities,sensor devices collect and release all kinds of urban data all the time.These data may become abnormal due to environmental interference,equipment failure or human tampering.Aiming at the problem of anomaly data filtering,a set of extraction protocols suitable for heterogeneous data was defined.Based on the extraction protocol,a distributed anomaly data classification filtering algorithm was designed based on Gaussian membership.Simulation experiments verified that the proposed algorithm could accurately filter abnormal data,so as to achieve the purpose of improving urban computing efficiency and reducing network overhead.

Key words: urban computing, Internet of things, big data, abnomaly filtering, mobile edge computing