Nonstationary Velocity Estimation

.4 nonstationary noise may frequently be approximated by the product of a stationary noise and a deterministic function of time. From observations of the su77i of such a nonstationary noise and a linear signal, an estimate of the rate of change of the signal is found. More exactly, a random function, x(t), is assumed to be one of the following: a + bt + g(t)n(t), a + bt + g(t) f h(t - r)n(r) dr or where a and I) are constants, h(t) is the impulse response of a lumped parameter filter, n(t) is white noise and g(t) is a nonzero deterministic function. A least squares estimate of b is found as a linear operation on a finite sample of x(t).