Citation
Abstract
We present an algorithm to losslessly compress a sequence of integers. The intended application is to encode seismic data at nodes in a network of seismometers. The algorithm performs predictive compression, using adaptive linear filtering to predict sample values and encoding variable numbers of samples into fixed-length packets. To accommodate packet losses, the packets include sufficient overhead data to ensure that samples in each packet can be decoded without requiring data from preceding packets. Compression results are presented for seismic test data sets.
Details
- Volume
- 42-173
- Published
- May 15, 2008
- Pages
- 1–13
- File Size
- 424.9 KB