We now generate

1.5

seconds of binary noise. In this noise process there are only two possible amplitudes at each time point: either +1 with probability p or –1 with probability 1–p. Further, we choose p = 1/2 and this implies the “fair coin” that we have looked at in Example 4.1 and Example 4.4.

1. Based upon the data shown in the bottom row, is it reasonable to conclude that b[n] is a sample of a white noise process? Explain your reasoning.


2. How do you perceive the loudness of b[n] compared to g[n] and u[n] from the two amplitude histograms and/or the signals themselves explain your answer?


3. What does this mean? Although we devote a good deal of time to examining the Gaussian noise process (and for good reason), is it the only noise process that can be described as white noise?


4. Hard: Are all stationary Gaussian noise processes white?

Proceeed to another part of this exercise by choosing the variant below.