Research On Real-time Data Acquisition And Compression System Of Telemetry Data Using LZW Algorithm

Huge amounts of data are generated while sampling various signals in telemetry systemand it is challenging for us to process all the data，also it is cost not only to transfer, but also tostorage the data. However, it has been demonstrated in statistics way that the redundancy rateof the data acquired and transmitted in a standard telemetry system is as high as90percent ormore. So the potential of telemetry data for being compressed is obvious, and there is strongdemand for a telemetry system with big size of data to use data compression technique toreduce signal space. Usually the data come from fast-changing parameters takes up a greaterportion as high as80of all the data in the domestic telemetry system. In consideration of this,there will have a significant effect on data space reduction and in this paper we put the focalpoint on real-time data acquisition and compression system of fast-changing parameters, andthe number of channel is made6.Firstly to guarantee the completeness of the important telemetry data here only losslesscompress algorithms are taken into account. Then after making comparison among severalwell known lossless compression algorithms such as the Huffman code，arithmetic code,dictionary-based code and so on in respect of most common factors to evaluate compressionperformance. The Lempel-Ziv-Welch (LZW) code is chosen for featuring in coding anddecoding fast, costing fewer, being simply implemented in hardware, high compression ratiorelatively, which meet the main need of the system we researched for its main characteristic,is real-time compression. Using LZW code, a scheme with platform implemented by FPGAdevice XC3S1400AN is proposed.Some solutions to optimize the flat LZW code are proposed with analysis made by toolsof Matlab, Modelsim combined with ISE, which covered essential aspects about dictionarysize and structure, table look-up method, dictionary upstate scheme and so on. Especially, amulti-hash table look-up method is presented which is proved by test has facilitated code speed fairly. The dictionary is simplified to save resource additionally, and the entries in thedictionary not referred in the recent past are eliminated, which result in higher compressionratio.To enhance its error resistance and reliability, Hamming code is implemented.Finally, a test is taken to detect the performance of the proposed real-time dataacquisition and compression system. Experimental results show the system work well asdesigned with compression ratio higher than1.8:1.