Publications Details

Real-time data compression using a FFT digital signal processor

Brady, E.

This report describes a hardware implementation of a fast fourier transform (FFT) based real time data compression system. The system is currently configured to compress and analyze airborne vehicle vibration data but it can be utilized for compressing any one- dimensional data stream. The algorithm uses a two-stage process to compress successive stationary time periods of input data. The first stage compresses data by transforming continuous blocks of 2048 points to the frequency domain using Welch's method. The second stage provides further compression by logarithmically averaging adjacent frequency bins of the transformed signal. These compression techniques result in a bandwidth savings from 8:1 to 64:1 per channel compared to strictly analog or digital transmission techniques. The complete technique as well as the hardware used to implement it is described in detail in this report.