The Efficiency of Detection Algorithms for Optimal FTN Signals

Abstract

Optimal signals represent the branch of FTN signaling and may be used for spectral efficiency improvement. Such signals may be detected with a simple coherent bit-by-bit algorithm if the transmission rate is equal to the one defined in the optimization problem. In this case, bit-error rate (BER) performance is close to the theoretical one. Extra increase in symbol rate leads to the necessity of application of more complex detection algorithms. In this paper, different detection algorithm including Viterbi algorithm and maximum likelihood sequence estimations are compared. It is shown that for an 11% increase in the symbol rate algorithm of partial enumeration with 2 iterations may be used instead of the Viterbi algorithm because provides almost the same BER performance with less computational complexity.