Self-adaptive maximum-likelihood sequence estimation

Many problems in digital communications can be modeled by means of a discrete-time finite-state Markov process representing the signal which is observed in independent identically distributed noise. If the parameters of the process are known, the problem is well understood and the optimum solution to the problem is to determine the state sequence which is most likely in light of the received data. This approach is referred to as maximum likelihood sequence estimation (MLSE) and can be performed computationally efficiently using the Viterbi algorithm. The present authors consider the important case when some or all of the process parameters are unknown. The traditional approach to this problem involves embedding a known “training sequence” in the data and estimating the parameters from the resulting received signal. Then, the data are extracted using the estimated parameters. In contrast, the authors propose to exploit the structure and finiteness of the state space of the signal to determine the most likely state sequence without resorting to a known training sequence. They refer to this approach as self-adaptive MLSE