Hi,
I have a mixed signal which consists of three individual signals. I would like to get back the original three signals from the mixed signal. The procedures I used in MATLAB are:
1. Applied FFT on the mixed signal
2. Used lowpass and bandpass filter to separate the signals
3. Used filtfilt (https://www.mathworks.com/help/signal/ref/filtfilt.html ) to get back the original signals.
However, I'm not getting the reconstructed signals exact/close to the original signals. The length of L1 and L2 should be same/very close for both original signal and reconstructed signals (see Fig. 5)
Any help/suggestions would be highly appreciated.
Regards,
Sohel
Hi , don't use these techniques .... there are several mathematical singnal processing techniques like VMD ,EEMD,EMD,ICA blind source separation method, gram's decomposition method to extract those three individual signal from the mixed one....I will recommend you to use VMD ....it is available in MATLAB ....
or use EEMD .....WHCH GIVES VERY GOOD RESULT IN MODE Decomposition ie individual signal extraction from mixed signal.....already I have worked in EEMD ..... With a published paper on EEMD
you can refer to that paper
https://www.researchgate.net/publication/347759876_i-PRExT_Photoplethysmography_Derived_Respiration_Signal_Extraction_and_Respiratory_Rate_Tracking_Using_Neural_Networks
Though I have the code of ICA ,EEMD ,VMD from scratch
- Badges
- Science topic
- Similar topics
- Biological Science
- Zoology
- Invertebrate Zoology
- Arthropods
- Insects