Here are two figures from the same signal, the 2nd one(B) is after digital predistortion at transmitter end. How can you describe it? What are the improvements and advantages of this technique?
Much appreciate your answers.
Signal Processing Stack Exchange is a question and answer site for practitioners of the art and science of signal, image and video processing. It only takes a minute to sign up.Sign up to join this community
You can compare levels (for instance the upper envelope in decibels) globally, by frequency range, their dynamics (variations across frequencies, like slopes) per range, and the transitions between frequency ranges, to start with.
Here for instance, the peak SNR or the span of the enveloppes (thickness of the black parts) are similar. However, the right-hand side spectrum is flatter in both parts. Possibly, the response is "more even" on the low frequency range (0 to 0.5 Hz times a cropped power of ten) and the higher one. This could be good on the signal part, maybe not so on the right end of the spectrum (the noise is 5 dB higher at 1.2 Hz times a cropped power of ten). The steepness between those parts some faster and the "step" is a bit bigger (-20 dB on the left, perhaps -25 dB) of the right.
Those morphological operations should then be matched with the purpose of the device which is meant to improve something, possibly.