In this paper we first describe a Maximum A Posterior (MAP) based
sequence estimation approach for unknown, fast fading, frequency
selective digital communications channels. The approach
incorporates prior probabilistic knowledge of the channel
via a stochastic channel model. We then assume a first order Gauss-Markov
channel model to derive a specific MAP estimator, and we describe a
Per Survivor Processing (PSP) algorithm as a realistic
approximation of the optimum estimator. Monte Carlo simulations results
are then presented: first to illustrate the performance realized when
incorporating prior channel knowledge and its dependence on channel
fading rate; and second to study the sensitivity
of this MAP estimator to inaccuracies in the assumed values of parameters
of the Gauss-Markov model.
This research was supported by ONR under Grant N00014-98-1-0892.
