# Negative instantaneous frequency in Hilbert Huang transform

How can I get rid of negative instantaneous frequencies (IF) of Hilbert Huang transform. Negative IF is physically meaningless, so why they are always exist? How they should be interpreted? I use MATLAB emd function to decompose a signal to its intrinsic mode functions (IMF). Then I employ instfreq function of MATLAB to evaluate IF. Unfortunately always IF has some negative values. There is a suggestion in following link and it is claimed that Dr. Huang has proposed it: WHY DO WE HAVE NEGATIVE FREQUENCY ALL THE TIME?

• Welcome to SE.SP! Unless your signal is monocomponent (or satisfies other conditions), you will almost always get negative instantaneous frequencies. See Huang et al. – Peter K. Mar 8 at 20:58