面向动态本地无线环境的电波传播特征认知方法研究

Based on the principle of user-centric and goal-driven which is usually applied by cognitive radio network, the project intend to improve the ability of accurate constructing the local wireless environment map under arbitrary scenarios around the end user. To this end, we would at first make comprehensive understanding of the multi-dimension and multi-scale characteristics of the dynamic wireless channel from the observed data, then inspired by the idea of dynamic map construction in machine vision, we apply a hybrid modeling method to combine the complex radio propagation mechanisms with automatic information fusion method, based on multiply continuous radio snapshots obtained by the end user, so that a novel radio propagation prediction method will be realized, which can be cognitive to the local radio environment. This method can enable fast tracking of the space-time change in radio propagation for local radio environment, and improve the end user's multi-dimension cognitive capability for dynamic and complicated heterogeneous network environment, making self-adaption capability of communication service at the end user implemented. This research achievement can perfect both radio frequency environment cognitive and location awareness theories, and further facilitate the solution of some key problems in communication quality evaluation method under heterogeneous and ubiquitous network. Besides, a novel way for applying the radio propagation to engineering application in cognitive radio network can also be provided by this method.

基于认知无线网络以用户为中心、以目标驱动为框架的理念，本项目旨在提高用户终端在任意场景中精确构建本地无线环境地图的能力，从认知终端可观测的角度深入理解动态无线信道的多维和多尺度特征，借助机器视觉中动态地图构建的想法，根据用户终端采集的多个连续无线电快照，采用机理与数据的混合建模方法，将复杂电波传播机制和自主信息融合方法进行有效结合，实现具有本地无线环境认知能力的电波传播预测方法，使之具备多维、多尺度和动态的预测特征，进而能够快速跟踪无线环境电波传播特征的空时变化，提升认知终端对于动态复杂异构网络环境的多维感知能力，完成认知终端的通信服务自适应。该研究成果可进一步完善无线频谱环境认知以及位置感知理论，从而为无线网络在异构化和泛在化趋势下的通信质量评估方法提供有效解决途径，并为电波传播在认知无线网络中的工程应用提供一种新颖的思路。

基于认知无线网络以用户为中心、以目标驱动为框架的理念，本项目旨在提高用户终端在任意场景中精确构建本地无线环境地图的能力，从认知终端可观测的角度深入理解动态无线信道的多维和多尺度特征，借助机器视觉中动态地图构建的想法，根据用户终端采集的多个连续无线电快照，采用机理与数据的混合建模方法，将复杂电波传播机制和自主信息融合方法进行有效结合，实现了具有本地无线环境认知能力的电波传播预测方法，使之具备多维、多尺度和动态的预测特征，进而能够快速跟踪无线环境电波传播特征的空时变化，提升了认知终端对于动态复杂异构网络环境的多维感知能力，完成了认知终端的通信服务自适应。根据项目研究成果，在国内外重要期刊或会议上发表学术论文42篇，申请国家专利7项。