A channel H(z) is followed by a TEQ A(z), and the residual output B(z) is to be chosen so that its length is less than that of the cyclic prefix. Each signal is expressed as a polynomial, with varying degree, and with corresponding coefficients. Once expanded, the coefficients of similar degree can be equated. The error signal can then be determined in terms of coefficients corresponding to the TEQ and the residual signal. The coefficients of the channel response can be derived from the channel training estimates. The error signal is minimized and the result is solved for in terms of the desired TEQ coefficients.
展开▼
