A sensor node compression algorithm based on H.264 integer DCT transform in WSN

Xuejun Ren,Dingyi Fang

DOI: https://doi.org/10.1049/cp.2010.1056

2010-01-01

Abstract:Efficient utilization of energy is a core area of research in wireless sensor networks since sensor nodes are powered by batteries which cannot be generally changed or recharged. Most of the energy is spent in communication module for data transmission. Data compression methods to reduce the number of bits to be transmitted by the communication module will significantly reduce the energy requirement and increase the lifetime of the sensor node. The spatio-temporal correlations among sensor data are a significant and unique characteristic of the WSN. The discrete cosine transform can significantly reduce these correlations due to its energy compaction property. However, the original DCT transform algorithm is too complicated to be applied in the sensor nodes. This paper proposes a lossless algorithm based on the tailor of H.264 integer DCT Transform. The algorithm transforms sensor data to result data which are input to an entropy encoder. The entropy encoder performs compression by encoding result data more compactly based on their statistical characteristics. Compared with the typical lossless algorithm, the purposed algorithm indicated better compression for nonslowly- varying data.