Lifting Wavelet Compression Based Data Aggregation in Big Data Wireless Sensor Networks

Abstract

The redundancy of sensing data in wireless sensor networks (WSNs) gives rise to longer transmission delays and more energy consumption. In this paper, we focus on the energy-efficient data redundancy elimination and compression with the objective of recovering the original data. To balance aggregation load of a large-scale WSN, we propose a novel energy-efficient dynamic clustering algorithm by utilizing spatial correlation, which can achieve a distributed compressive data aggregation in each cluster head. Furthermore, we propose a distributed fast data compression approach based on eliminable lifting wavelet to reduce the amount of raw data. Also, it offers high fidelity recovery for the raw data. Extensive experimental results demonstrate that our clustering method based on data correlation clustering (CDSC) for data aggregation outperforms other methods on prolonging network lifetime and reducing the amount of data transmitted. In particular, our data compression aggregation algorithm can achieve 98.4% recovery accuracy when the compression ratio equals 1.3333.