Continuous Channels and Gaussian Channels
The BSC and BEC models, inputs are binary {0 , 1} and it has the practical implementations and operations observed in most digital communication systems. But if inputs are continuous and distributed with mean 0 and some variance and output are also continuous then the channel is called Gaussian Channel.
Gaussian channel represents the most fundamental form of all types of communication channel systems and is used to provide meaningful insights and theoretical results on the information carrying capacity of channels.

That is, the channel output is troubled by additive white Gaussian noise (AWGN).
Let X and Y be two random variables with joint density f(X=x, Y= y) and their marginal densities are f(X = x) f(Y = y). The mutual information, I(X , Y) provides a measure of the amount of information that can be carried by the continuous channel, then the mutual information between X and Y is defined as:

No More
Statlearner
Statlearner