Gaussian Particle Filtering Approach for Carrier Frequency Offset Estimation in OFDM Systems

Abstract

We propose Gaussian particle filtering (PF) approach for estimating carrier frequency offset (CFO) in OFDM systems. PF is more powerful especially for nonlinear problems where classical approaches (e.g., maximum likelihood estimators) may not show optimal performance. Standard PF undergoes the particle impoverishment (PI) problem resulting from resampling process for this static parameter (i.e., CFO) estimation. Gaussian PF (GPF) avoids the PI problem because resampling process is not needed in the algorithm. We show that GPF outperforms current approaches in this nonlinear estimation problem.