A Classification of Lossless and Lossy Data Compression Schemes

Abstract

Data compression is a promising scheme to increase memory system capacity, performance and energy advantages. The compression performance could affect the overall network performance when compression scheme is implemented in a communication field. Many data compression schemes have been introduced. Most of other researchers choose very limited parameters to analyze the performance of the selected data compression scheme. This paper classifies the major data compression schemes according to nine different perspectives, such as homogeneity, purpose, accuracy, structuring of the data, repetition distance, structure sharing, number of passes, sampling frequency, and sample size ratio. Various data compression schemes are examined and classified according to the parameters mentioned above. The classification will provide researchers with the in-depth insight on the potential role of compression schemes in memory components and network performance of future extreme-scale systems.