Delay and Doppler estimation by time-space partition of the array data

An instrumentationally attractive technique for estimating the Doppler time-compression and its time derivatives is proposed and analyzed in this paper. The basic scheme consists of (M - 1) generalized cross correlators (GCC's) in parallel (where M is the number of sensors in the receiving array), each of which is implemented successively to generate a sequence of differential delay measurements. The resulting spatial-temporal differential delay data are then compressed into the estimates of the Doppler and its derivatives using a least squares (LS) filter. It is shown that for signal-to-noise ratio (SNR) conditions above the threshold level, the indicated two-step estimation procedure is very nearly optimal. When operating near the threshold point, however, the proposed scheme is only distinctly suboptimal.