Cross covariance of x and y, in statistics, is a measure of the similarity between x and shifted versions of y, as a function of the shift (lag). The cross covariance is given by the equation
where E is the expectation operator, and the processes have mean functions vt=E[Yt] and μt=E[Xt]
In signal processing, cross-covariance has a slightly different definition: it measures the similarity between two signals, and is a function of the time between signals. It is sometimes called the sliding dot product or the cross-correlation.
Note that, in statistics, the cross-correlation and cross covariance are related but are not the same thing.