Joint epoch estimation for multiple FH signals

Joint hop-timing (epoch) estimation for a desired frequency-hopping (FH) signal observed in the presence of other interfering FH signals is addressed. A coarsely channelized preprocessor is used to suppress the frequency and phase dependence of the signals. Coarse channelization implies a bank of filters, much smaller in number than the size of the candidate hop frequency set, which covers the total bandwidth shared by all of the signals. It is assumed that the signals all have the same hopping rate and set of candidate hop frequencies, but have distinct power and timing epochs. Average likelihood functional does not yield an easy closed form expression; therefore, the generalized likelihood concept is used. It is shown that the solution can be efficiently implemented using trellis search algorithms coupled with recursive parameter estimators.