无线通信网络为智慧城市万物互联提供了最坚实的网络支撑，而频谱资源则是其核心资源之一。为了充分掌握城市中频谱资源分布状况，需要对频谱信号进行准确的建模与分析，并以此为据进一步对资源进行高效应用。本项目通过群智感知的方式对城市海量时空频谱数据进行采集，分析其在时间、空间上的分布特点，研究数据不均匀与数据缺失问题，探究数据补全方法，对城市中无线频谱资源进行准确建模与预测，挖掘其有效信息，进而建立频谱地图，实现在城市精准定位等应用中。并将研究成果应用在城市管理中，通过融合海量多源异构数据，对城市人口流动进行精准建模，优化传染病传播模型，寻找密切接触人群，优化城市传染病精准防控策略。

Wireless communication network provides the most fundamental support for the mass connections in the smart city, where spectrum is the core resource. Thus it’s very important to apply accurate spectrum modeling and analysis in the city, in order to grasp the spectrum resource distribution, and then improve the resource utilization. Based on the crowd sourcing magnanimity spatial-temporal spectrum data collected in the city, this project proposes to analyze the spacial and the temporal distribution of the spectrum data, propose the data completion methods, then conduct precise modeling and forecasting. Besides, we build the fine-grained spectrum map to perform high precision city localization. Moreover, we apply the results on the city management. By conducting multi-source cross-domain heterogeneous data fusion, we model the population movement and the epidemic propagation model, search for the close contact, and propose a fine-grained epidemic controlling strategies.