m (Protected "Text Slecture1" ([Edit=Allow only administrators] (indefinite) [Move=Allow only administrators] (indefinite))) |
|||
Line 30: | Line 30: | ||
If you have any questions, comments, etc. please post them on [[Downsampling_Soonho_Kwon_slecture_review|this page]]. | If you have any questions, comments, etc. please post them on [[Downsampling_Soonho_Kwon_slecture_review|this page]]. | ||
+ | ---- | ||
+ | [[2014_Fall_ECE_438_Boutin_digital_signal_processing_slectures|Back to ECE438 slectures, Fall 2014]] |
Latest revision as of 09:00, 14 March 2015
Contents
- 1 1. Introduction
- 1.1 Sampling is essential to convert analog signal to digital signal. In this page, I would specificallyexplain about downsampling. In digital signal processing, the decimator performs decimation, which isdownsampling a signal. In other words, when a digital signal is downsampled, the signal's samplingrate would be reduced.
- 1.2
- 2
- 3 2. Derivation
- 4
- 5 3. Example
- 6 If) D = 2
- 7
- 8
- 9 4. Conclusion
1. Introduction
Sampling is essential to convert analog signal to digital signal. In this page, I would specifically
explain about downsampling. In digital signal processing, the decimator performs decimation, which is
downsampling a signal. In other words, when a digital signal is downsampled, the signal's sampling
rate would be reduced.
2. Derivation
3. Example
If) D = 2
4. Conclusion
Downsampling reduces the magnitude by a factor of D, and increases the width of signals by D in the frequency domain. Moreover, it repeats it's signals every 2π and to prevent aliasing, decimator needs a lowpass filter before the downsampling.
Questions and comments
If you have any questions, comments, etc. please post them on this page.