A Sensor Node Lossless Compression Algorithm Tailored from H.264 integer DCT Transform

摘要

efficient utilization of energy is a core area of research in wireless sensor networks. 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 discrete cosine transform can significantly reduce the spatio-temporal correlations among input data and then achieve effective compression ratio. However, the original DCT transform algorithm is too complicated to be applied in the sensor nodes. This paper proposes a lossless algorithm, which tailored from H.264 integer DCT transform, particularly suited for memory and computational resource constrained wireless sensor node. 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 non-slowly-varying data.