Assigning parameter estimates for N-survivor processing

In an attempt to improve N-survivor processing (NSP), which lies between pe r-survivor processing and tentative decisions, a new algorithm is proposed for assigning parameter estimates to survivors. The proposed algorithm cons iders the history of the survivors in a trellis diagram by evaluating the d istance between the hypothesised input vectors that contain the path histor y. It will be shown through computer simulation that NSP employing the prop osed algorithm can outperform the original version.