IEEE Signal Processing Letters, cilt.30, ss.603-607, 2023 (SCI-Expanded)
The problem of detecting the state sequence of a binary stochastic process with multiple sensors is considered. It is assumed that the sensors' observations are coupled both spatially and temporally. Copula theory is used to model the unknown spatial coupling across the observations of different sensors while the temporal dependency is taken into account using a first order Markov chain. A per-survivor processing-based algorithm is proposed to estimate the unknown parameter vector and determine the correct state sequence with reduced computational complexity. Numerical examples indicate that it is possible to keep track of the underlying state sequence while satisfactorily estimating the unknown parameters with the proposed method.