基于车联网的数据动态压缩采集策略

Dynamic Data Compression Strategy Based on Internet of Vehicle

车联网能高效地实现感知区域的覆盖,因此被应用于大规模城市感知。同时,为了解决车联网难以传输大量数据的问题,一些研究者使用压缩感知对具有时空相关性的数据进行压缩。但是,目前在车联网中应用压缩感知的研究并没有考虑数据和车辆分布变化的特性,很可能导致不可接受的误差。为了保证数据的重构精度,提出面向车联网的动态压缩感知方法。该方法能自动分析感知对象的数据特征、车辆分布和观测数量之间的关系。在压缩感知的基础上加入观测数量调整功能,通过对当前感知对象的数据特征和车辆分布的分析,实时调整压缩感知中观测矩阵的参数,从而控制观测数据的数量,提升重构精度,实现更高质量的数据传输。实验表明与现有车联网中的压缩感知方法相比,面向车联网的动态压缩感知方法在重构精度上提升了15.3%。

Internet of Vehicle(IoV)can effectively cover the sensing area,so it is applied to large-scale urban sensing.Meanwhile,in order to solve the problem that it is difficult for the IoV to transfer a large amount of data,the compressive sensing(CS)is used to compress the data with spatio-temporal correlation by some researchers.However,the current researches on the applications of CS in the IoV do not consider dynamic changes in data characteristic and vehicle distribution,which may lead to unacceptable errors.In order to ensure the accuracy of data reconstruction,this paper proposed a dynamic CS approach in the IoV.The approach can automatically analyze the relationship among data chara-cteristic,vehicle distribution and the number of measurements.Then,based on the CS,a function of adjusting the number of measurements is added.Through the analysis of data characteristic and vehicle distribution,the parameters of the observation matrix in CS is adjusted in real time so as to improve the accuracy of reconstruction to achieve higher quality data transmission.The experiment shows that the proposed dynamic CS method improves the reconstruction accuracy by 15.3%compared with the existing CS method in the IoV.