Cyclic Autocorrelation Function.

Equation (9.7) involves an aperiodic correlation of data ${\bf x}$. The extension to cyclic correlation estimates can also be formulated in terms of quadratic forms by wrapping each of the diagonals. For example, for $N=6$, ${\bf P}_2$ becomes

$${\bf P}_2 = \frac{1}{2N} \left[ \begin{array}{cccccc} 0 & 0 & 1 &... ... 0 & 1\\ 1 & 0 & 1 & 0 & 0 & 0\\ 0 & 1 & 0 & 1 & 0 & 0\\ \end{array}\right].$$ (9.8)