Joint CFO Estimation and Symbol Detection for Embedded Pilot in MIMO-OFDM

Abstract

This work introduces a novel source separation and Carrier Frequency Offset (CFO) estimation method based on a semi-blind multi-modulus approach with embedded pilots for Multiple-Input Multiple-Output Orthogonal Frequency Division Multiplexing (MIMO-OFDM) communication systems. A hybrid cost function combining Least Squares (LS) and Multi-Modulus (MM) criteria is devised for semi-blind source recovery and CFO estimation. To minimize this cost function, an efficient block gradient descent algorithm is employed, along with an optimized step-size strategy to speed up convergence. Furthermore, a low-complexity adaptive version of the algorithm is proposed for joint equalization and CFO estimation in slow-varying channels. Simulation results demonstrate that the proposed method enhances spectral efficiency while maintaining low computational complexity and achieving enhanced convergence.