Analysis and Comparison of Permutation Entropy, Approximate Entropy and Sample Entropy

Permutation entropy, approximate entropy and sample entropy can be used to measure the complexity of time series. The sinusoidal signal changed with length, amplitude, phase and frequency is measured by these three kinds of entropy. Simulation results show that the entropy value is stable when the data length is longer than some value, for example, 2048. It is also shows that the change of amplitude and phase does not affect the entropy value. The permutation entropy can better describe the frequency change of the signal.