Compressing Computer Network Measurements Using Embedded Zerotree Wavelets

Monitoring and measuring various metrics of high data rate and high capacity networks produces a vast amount of information over a long period of time. Characteristics such as throughput and delay are derived from packet level information and can be represented as time series signals. This paper looks at the Embedded Zero Tree algorithm, proposed by Shapiro, in order to compress computer network delay and throughput measurements while preserving the quality of interesting features and controlling the level of quality of the compressed signal. The quality characteristics that are examined are the preservation of the mean square error (MSE), the standard deviation, the general visual quality (the PSNR) and the scaling behavior. Experimental results are obtained to evaluate the behaviour of the algorithm on delay and data rate signals. Finally, a comparison of compression performance is presented against the lossless tool bzip2.