Research On ARM-processor Based Data Compression And Storage

With the rapid development of digital technology and the further popularization ofinformation technology, The amount of data requiring transmitting, storing, andprocessing increases sharply. These give a huge pressure to bandwidth, storagecapacity and processing speed, and require new approach of data compression.This research is based on the actual project titled endoscope cleaning anddisinfection program. With the standardization of requirements with respect toendoscope cleaning system in major hospitals, a systematic, human-friendlyprocedure of endoscope cleaning and disinfection machine has been graduallyimproved. The records of the procedure are changing from manually recording todigitally recording. This increases the pressure of data collecting, transmitting andstoring, especially for the ARM processor, which is at the bottom of the dataprocessing architecture. In case of communication outage between the ARMprocessor and the server, the accumulation of data brings a severe test to the ARMprocessor, which is only256Mb. Thus, this thesis will research on ARM-processorbased data compression.This research aims at alleviating the pressure on the ARM processor. Afteranalyzing the actual data, we summarized the format of the packet and recognized thecharacteristic of finite state set of data. On the basis of analyzing the vectorquantization theory and LZ coding algorithm, we propose a new improved datacompression algorithm, which is suitable for finite state set data. The new algorithm isnamed Vector-searching based Self-adaptive data Compression Algorithm, VBSDCalgorithm for short.The new algorithm simplifies the procedure of dividing quantitative range andfinding quantized value in vector quantization principle, reduces the size of dictionaryin LZ coding algorithm, makes use of the idea of codebook in the vector quantizationand the ideas of dictionary lookup and adaptive coding in the LZ coding algorithm.Combined with characteristics of Nand Flash memory in the ARM processor, weultimately release the memory pressure of ARM processor, save storage space, extend the lifetime of Flash, and expand the storage capacity.Finally, we test the new algorithm in the real project. Based on the analysis of theability of compression and the quality of decoding, we conclude that this newalgorithm is stable and reliable.