Compressed sensing Kalman filter estimation scheme for MIMO system under phase noise problem

• Published in: IET communications 2020-12, Vol.14 (22), p.4108-4115
• Main Author: El-Mashed, Mohamed G
• Format: Article
• Language: English
• Subjects: Bayes methods; Bayesian Cramér–Rao lower bound scheme; common oscillators; compressed sensing; data‐aided least square estimator; error statistics; estimation schemes; high‐speed communication systems; independent oscillators; joint estimation; Kalman filters; KF‐based CS scheme; maximum likelihood estimation; mean square error methods; MIMO communication; MIMO system; multiinput–multioutput systems; phase noise; phase noise parameters; phase noise problem; phase parameters; Research Article; sensing Kalman filter estimation scheme; system parameters
• ISSN: 1751-8628; 1751-8636
• DOI: 10.1049/iet-com.2019.1272

Phase noise problem in oscillators can degrade the performance of high-speed communication systems. The author analysed the impact of phase noise problem on multi-input–multi-output (MIMO) systems under common and independent oscillators. The estimation of system parameters (i.e. phase noise and channel gains) is a challenging task. In this study, a data-aided least square estimator based compressed sensing Kalman filter (KF)-based compressed sensing (CS) scheme is proposed for tracking phase parameters. The signal model and estimation problem for the system are mathematically derived. Also, Bayesian Cramér–Rao lower bound (BCRLB) scheme is also derived. For joint estimation, the mean square error (MSE) and bit error rate (BER) performances of the BCRLBs and proposed scheme are compared. Results demonstrate that the proposed KF-based CS scheme gives low BER values and better performance compared to other estimation schemes. The utilisation of the proposed scheme helps in reducing the MSE of the MIMO system. Finally, the proposed scheme enhances the estimation of phase noise parameters for the MIMO system.