云存储环境中传感数据的压缩存储处理研究

发布时间：2018-07-01 17:47

本文选题：传感数据 + 压缩算法　；参考：《南京理工大学》2012年硕士论文

【摘要】：无线传感器网络节点生成大量冗余数据,这些数据在节点间的转发会引发一系列问题,如节点上有限能量的大量浪费、网络传输延迟、网络中海量传感数据存储处理困难等。这些问题都严重制约了无线传感器网络应用的进一步发展和大规模推广使用。针对这些问题,本文从以下两个方面来进行云存储环境中传感数据压缩存储研究：传感器节点数据压缩和海量传感数据压缩存储。一、考虑到单个节点传感数据间存在时空相关性,本文提出了分段常量近似与小波压缩相结合的PCA_W压缩算法,在误差可调的情况下针对该类时空相关的传感数据进行压缩处理。实验分析比较了PCA_W算法与其它两种压缩算法在数据重构误差、数据压缩比、压缩耗时和解压耗时方面的表现。结果表明PCA W算法可以显著减少冗余数据,有较高的压缩比并可以保证数据重构精度。二、考虑到网络中海量传感数据存在大量重叠、存储处理困难等问题,本文提出基于云存储的海量传感数据存储模型,使得上层应用数据存取访问更便捷、更有效；设计基于哈夫曼算法的海量传感数据压缩的方法,压缩海量传感数据,可以节约存储空间和带宽。实验验证了本文所提出的压缩方法在压缩比、压缩处理耗时以及解压处理耗时方面具有的优势。
[Abstract]:Wireless sensor network nodes generate a large number of redundant data, which will lead to a series of problems among nodes, such as a large amount of waste of limited energy on the nodes, network transmission delay, and the difficulty of storing and processing massive sensor data in the network. These problems have seriously restricted the further development of wireless sensor network applications and large-scale popularization. In order to solve these problems, this paper studies sensor data compression storage in cloud storage environment from the following two aspects: sensor node data compression and mass sensor data compression storage. Firstly, considering the spatio-temporal correlation between sensor data of single node, a PCAW compression algorithm combining piecewise constant approximation and wavelet compression is proposed, which can compress the sensing data with adjustable error. The performance of PCAW algorithm and other two compression algorithms in data reconstruction error, data compression ratio, compression time and decompression time are analyzed and compared experimentally. The results show that PCA W algorithm can significantly reduce the redundant data, have a high compression ratio and ensure the accuracy of data reconstruction. Second, considering that there is a lot of overlap in the massive sensing data in the network, the storage model of the mass sensing data based on cloud storage is proposed in this paper, which makes the access to the upper application data more convenient and effective. The method of mass sensing data compression based on Huffman algorithm is designed. The storage space and bandwidth can be saved by compressing the mass sensing data. Experimental results show that the proposed compression method has advantages in compression ratio, compression processing time and decompression processing time.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP333

【参考文献】