Processing Technique
A digital signal processing operation reduces the sampling rate of a digitized signal while preventing aliasing of high-frequency noise. The operation, termed decimation filtering, combines a low-pass filter with downsampling to eliminate unwanted signal components. This step preserves the integrity of the remaining band of interest.
Signal Aliasing
Downsampling without prior filtering folds high-frequency noise directly into the primary signal band. The decimation filtering process removes these out-of-band components before the sample rate reduction occurs, ensuring that the digitized signal remains free of spurious spectral lines.
Filter Implementation
Most digital systems utilize finite impulse response filters to achieve a linear phase response during the sampling rate reduction. In delta-sigma analog-to-digital converters, decimation filtering is executed in multiple stages to optimize the computational load on the processor. This multi-stage approach reduces the number of mathematical operations by performing the initial filtering at the highest rate with simple coefficient structures.
Performance Tradeoff
Selecting a narrow transition band in the filter requires a higher filter order, which increases the processing latency of the system. Systems that require real-time feedback must balance the suppression of out-of-band noise against this latency. The decimation filtering parameters are selected to meet the specific requirements of the control loop.