WLC42-1: Multiple-Model Rao-Blackwellized Gauss-Hermite Filter for Joint Frequency Offset and Channel Estimator for the Semi-Blind MIMO-OFDM Receiver

In this paper, we propose a new statistical joint frequency offset and channel estimation for the blind MIMO-OFDM system, where a state and parameters of dynamic systems are unknown. To reduce a channel mismatch problem, we use the fixed structure interacting multiple model (FS-IMM) and the Rao-Blackwellized Gauss-Hermite filter to estimate a model-dependent channel and offset parameter, which enters an observation function in a nonlinear manner. The resulting structure consists of M parallel bank of Kalman filters driven by sampled nonlinear state variable sequences.