Publications

A common limitation to performance in data acquisition systems is storage of the collected data. Compressing the data would increase the amount of data that could be stored. However, most compression routines require that the data be collected and analyzed before compression is performed. Also, these compression routines often store the information required for decompression along with the data, thus decreasing the storage available for data. One solution to this problem is to create an encoding tree known to both the encoder and the decoder based on apriori knowledge of the data. Once the tree is created, optimal encoding schemes such as the Huffman algorithm may be used on the data as it is being collected. In this way the data is compressed as each byte is received and there is no overhead associated with storing decompression data. In this paper the idea of using a fixed Huffman tree is explored and the results are compared to a defacto standard in data compression, PKZIP.

In RF telemetry (TM) the allowable RF bandwidth limits the amount of data in the telemetered data set. Typically the data set is less than ideal to accommodate all aspects of a test. In the case of diagnostic data, the compromise often leaves insufficient diagnostic data when problems occur. As a solution, intelligence was designed into a TM, allowing it to adapt to changing data requirements. To minimize the computational requirements for an intelligent TM, a fuzzy logic inference engine was developed. This reference engine was simulated on a PC and then loaded into a TM hardware package for final testing.