An improved algorithm for blind equalization based on the tricepstrum

In this paper, we give an improved algorithm for blind equalization based on the tricepstrum. The tricepstrum equalization algorithm (TEA) proposed by D. Hatzinakos and C.L. Nikias (1989) has more excess MSE after convergence and it may cause error determination at the system output. For decreasing the excess MSE, we give a new cost function by combining the TEA algorithm and Sato algorithm. The results of theoretic analysis and simulation show that our algorithm has good convergence performance and can overcome the intersymbol interference (ISI), moreover it has lower excess MSE.