期刊名称：International Journal of Electronics Communication and Computer Engineering
印刷版ISSN：2249-071X
电子版ISSN：2278-4209
出版年度：2013
卷号：4
期号：5
页码：1341-1348
出版社：IJECCE
摘要：Vibration tests on satellites generate huge amount of data of the order of few Giga bytes per test and about tens of Terra bytes of data every year. This data represents vibration levels monitored at various locations of the satellite during the test. The files storing this huge data are archived in the data server computer system. Presently, it is not practical to access this mammoth data over intranet for post processing the data at a later time. Hence, the post-processing of vibration data encounters two challenges: archival of huge data in server system and accessing the mammoth data over intranet backbone of the organization. In the present work, an efficient compression algorithm using unique wavelets, proposed to be named as randlet, sinelet, and shocklet. These wavelets are designed using perfect reconstruction filter banks, to compress the satellite vibration test data which reduces the storage requirement and access time for retrieval of the data from the server system. A comparison is made between the algorithms using standard transformations such as Discrete Cosine Transform, Fast Fourier Transform, Haar Wavelet Transformation and the transformation using the proposed wavelets. It is concluded in this paper that the proposed wavelets provide optimal performance in terms of compression ratio, reconstruction error compared to existing transformation techniques
