无线传感器网络中基于序列相关性的数据压缩算法

Data Compression Algorithm Based on Sequence Correlation for WSN

无线传感器网络(WSN)中传输的数据具有相关性和冗余性。如何有效降低网络中的数据量,延长网络生命周期,始终是WSN的研究热点之一。该文基于WSN中数据序列的相关性,提出一种两步数据压缩算法(TSC-SC)。网络中的簇首和簇内节点执行各自的压缩算法:簇首首先执行相关性分组算法,将数据分组,减少簇内节点的计算量以及消除簇内数据的空间相关性;簇内节点对多属性数据分类压缩,并将压缩参数传至簇首,簇首解压后再次进行分类压缩,进一步消除数据相关性,减少节点数据冗余度,降低通信能耗。为实现对压缩算法的综合性能评价,考虑基本的压缩要求和算法的计算能耗,提出了基于能量判别的算法评估模型(NCER)。仿真结果表明TSC-SC算法可以有效降低压缩比和压缩误差,充分减少数据传输量和网络的通信能耗,利用NCER指标能够直观地评价算法的性能。

The data has correlations and redundancy in Wireless Sensor Network（WSN）. How to reduce effectively the amount of communication data and extend the network life cycle is one of researching hot points. The Two-Step data Compression algorithm based on Sequence Correlation（TSC-SC） for WSN is proposed in this paper. The cluster head and the nodes in clusters perform different compression algorithms for themselves. In order to eliminate the spatial correlation of data and reduce the calculated amount, the cluster head nodes perform the grouping algorithm firstly, then the nodes in clusters perform the classifing compression to eliminate correlation for multi-attribute data, and pass the compression parameters to the cluster head; the cluster head perform the classifing compression again after decompressing the parameters. So the data-redundancy and communication energy consumption is further reduced. A new evaluation model named Network Compression Energy Ratio（NCER） based on energy discrimination is also proposed. The evaluation model realizes comprehensive evaluation of compression algorithms by considering both the basic requirements of compression and calculated energy consumption in the nodes. Simulation results show that TSC-SC algorithm can reduce the compression ratio and compression error effectively; the amount of communication data and energy consumption can achieve a satisfactory level in the network. The algorithm can be estimated directly using NCER.