Reconstructing Digital Signals Using Shannon's Sampling Theorem
Click name to view affiliation
Search for other papers by Joseph Hamill in
Current site
Google Scholar
PubMed
Search for other papers by Graham E. Caldwell in
Current site
Google Scholar
PubMed
Search for other papers by Timothy R. Derrick in
Current site
Google Scholar
PubMed
Researchers must be cognizant of the frequency content of analog signals that they are collecting. Knowing the frequency content allows the researcher to determine the minimum sampling frequency of the data (Nyquist critical frequency), ensuring that the digital data will have all of the frequency characteristics of the original signal. The Nyquist critical frequency is 2 times greater than the highest frequency in the signal. When sampled at a rate above the Nyquist, the digital data will contain all of the frequency characteristics of the original signal but may not present a correct time-series representation of the signal. In this paper, an algorithm known as Shannon's Sampling Theorem is presented that correctly reconstructs the time-series profile of any signal sampled above the Nyquist critical frequency. This method is superior to polynomial or spline interpolation techniques in that it can reconstruct peak values found in the original signal but missing from the sampled data time-series.
The authors are with the Biomechanics Laboratory, Department of Exercise Science, University of Massachusetts, Amherst, MA 01003.
© 2023 Human Kinetics