A parameter estimation algorithm for propagation channels based on two-layer evidence framework

In this contribution, a new algorithm derived based on a two-layer evidence framework is applied to estimating parameters in the generic stochastic channel models from measurement data. Different from conventional high-resolution parameter estimation algorithms, e.g. the space-alternating generalized expectation-maximization (SAGE), the method proposed is applicable to extracting both the parameters of multiple components in individual realizations of channel impulse responses and the statistical parameters of the wide-sense-stationary channel. Furthermore, the proposed two-layer evidence framework can be readily generalized to accommodate appropriate channel features of interest as certain prior information. Preliminary simulation results demonstrate the effectiveness of the proposed algorithm when being used to estimate the cumulative distribution function of delay spreads of propagation channels.