The improvement of turbo equalization through using turbo codes

In this paper, we propose an equalization algorithm based on serial concatenation turbo-like method-turbo code and parallel concatenation turbo-like method-turbo equalization to improve the BER performance of the general turbo equalization. Utilizing turbo code, we can get a more exact probability of channel encoder input data bits, needed in a maximum a-posteriori probability (MAP) decoder, to replace the independent and identically distributed (IID) assumption of this probability in general turbo equalization. From the relationship between the two turbo-like methods we can investigate the convergence property of turbo codes in turbo equalization, and then determine the appropriate degrees of iteration in the inner loop based on certain iteration degree in the outer loop.