I have been using this definition of autocorrelation very often Rx(t) = E[x(t)x(t+T)]
but suddenly I got stumped to find my textbook using this definition Rx(t) = Σx(t)x(t+T)
I could figure out expectation operation is nothing but averaging and so the first definition should lead to the second one,,, but in case if I am averaging N samples of x, where is the divided by N term or (1/N) in the second formula...