Abstract: An adaptive detector comprises an adaptive linear prediction filter (ALPF) nd a detection processor.The adaptive filter comprises an input filter, adapted to receive an input signal x(k) and delay it by a time .DELTA.. An L-point, or L-tap, adaptive filter has an input connected to the output of the input filter, a signal r(k) appearing at the output of the filter. A means for summing has two inputs, one for receiving the signal x(k) and the other for receiving an inverted signal from the adaptive filter, the output of the summer being an error signal .epsilon.(k). A feedback amplifier, whose input is connected to the output of the summer, takes a portion 2.mu., of the output signal .epsilon.(k) and feeds it back to the adaptive filter, thereby modifying the tap weights of the adaptive filter.The detection processor comprises a circuit, whose input is connected to the output of the adaptive filter, to receive the signal r(k), for performing a K-point discrete Fourier transform (DFT) on the signal r(k).

Abstract: An input signal X(j) is fed directly to the positive port of a summing function and is simultaneously fed through a parallel channel in which it is delayed, and passed through an adaptive linear transversal filter, the output being then subtracted from the instantaneous input signal X(j). The difference, X(j)-Y(j), between these two signals is the error signal .epsilon.(j). .epsilon.(j) is multiplied by a gain .mu. and fed back to the adaptive filter to readjust the weights of the filter. The weights of the filter are readjusted until .epsilon.(j) is minimized according to the recursive algorithm: ##EQU1## where the arrow above a term indicates that the term is a signal vector. Thus, when the means square error is minimized, W.sub.(j+1) =W.sub.(j), and the filter is stabilized.