[Paper Review] Phase Noise Estimation for Uncoded/Coded SISO and MIMO Systems

Arif Önder Isikman, Hani Mehrpouyan+1 more|arXiv (Cornell University)|Oct 23, 2012

Advanced Wireless Communication Techniques32 references3 citations

TL;DR

This thesis proposes low-complexity, iterative receiver algorithms based on the expectation-maximization (EM) framework and extended Kalman filtering for joint phase noise estimation and data detection in uncoded and coded SISO and MIMO systems. The key contribution is an EKFS-based EM algorithm that achieves near-perfect phase noise compensation at low-to-medium SNR, with performance close to ideal knowledge of phase noise, especially when using low-rate LDPC codes to improve soft decisions.

ABSTRACT

Non-ideal oscillators both at the transmitter and the receiver introduces time varying phase noise which interacts with the transmitted data in a non-linear fashion. Phase noise becomes a detrimental problem and needs to be estimated and compensated. In this thesis receiver algorithms are derived and evaluated to mitigate the effects of the phase noise in digital communication systems. In Chapter 3 phase noise estimation in single-input single-output (SISO) systems is investigated. First, a hard decision directed extended Kalman filter (EKF) is applied to an uncoded system. Next, an iterative receiver algorithm performing code-aided turbo synchronization is derived using the expectation maximization (EM) framework for a coded system. Two soft-decision directed estimators in the literature based on Kalman filtering are evaluated. Low density parity check (LDPC) codes are proposed to calculate marginal a posteriori probabilities and to construct soft decision symbols. Error rate performance of both estimators are compared through simulations. In Chapter 4 phase noise estimation in multi-input multi-output (MIMO) systems is investigated. First, a low complexity hard decision directed EKF is applied to an uncoded system. Next, a new receiver algorithm based on the EM framework for joint estimation and detection in coded MIMO systems is proposed. A low complexity soft decision directed extended Kalman filter and smoother (EKFS) that tracks the phase noise parameters over a frame is proposed in order to carry out the maximization step. The proposed EKFS based approach is combined with an iterative detector that utilizes bit interleaved coded modulation and employs LDPC codes. Finally, simulation results confirm that the error rate performance of the proposed EM-based approach is close to the scenario of perfect knowledge of phase noise at low-to-medium signal-to-noise ratios.

Motivation & Objective

• To address the detrimental impact of time-varying phase noise on high-throughput digital communication systems, especially in high-frequency microwave backhaul links.
• To develop low-complexity, iterative receiver algorithms that jointly estimate phase noise and detect transmitted symbols in SISO and MIMO systems.
• To evaluate and compare the performance of soft-decision directed estimators (KS-MLA and EKS) under varying system parameters such as block length and modulation order.
• To extend the EM-based framework to coded MIMO systems using bit-interleaved coded modulation and LDPC codes for improved reliability.
• To investigate the trade-offs between performance, complexity, and coding rate in phase noise compensation for practical system design.

Proposed method

• Derives a hard decision-directed extended Kalman filter (EKF) for phase noise tracking in uncoded SISO and MIMO systems.
• Proposes an iterative EM-based receiver for coded SISO systems, using the E-step to compute posterior probabilities and the M-step to update phase noise estimates via soft-decision directed EKFS.
• Employs LDPC codes to compute marginal a posteriori probabilities for soft decision symbols, enhancing estimation accuracy.
• Introduces a low-complexity soft-decision directed EKF and smoother (EKFS) for joint phase noise tracking and data detection in coded MIMO systems.
• Uses bit-interleaved coded modulation (BICM) with LDPC decoding to generate reliable soft decisions for iterative phase noise estimation.
• Employs computer simulations to evaluate bit error rate (BER), frame error rate (FER), and mean square error (MSE) of phase noise estimates across varying SNR and phase noise variance levels.

Experimental results

Research questions

• RQ1 How does the performance of soft-decision directed estimators (KS-MLA and EKS) depend on system parameters such as block length and modulation order?
• RQ2 Can an EM-based iterative receiver with EKFS achieve near-perfect phase noise compensation in coded SISO and MIMO systems?
• RQ3 What is the impact of coding rate and decoder iteration count on phase noise estimation accuracy and error rate performance?
• RQ4 How does the estimation accuracy of the proposed EKFS algorithm evolve across EM iterations, and what is the convergence behavior?
• RQ5 To what extent does phase noise estimation error affect the overall system performance, and how can it be mitigated through receiver design?

Key findings

• The EKFS-based EM algorithm achieves near-perfect phase noise compensation in coded MIMO systems at low-to-medium SNR, with performance close to the ideal case of perfect phase noise knowledge.
• For high phase noise innovation variance or long block lengths, the KS-MLA estimator degrades significantly due to unreliable soft decisions, while the EKS performs better under such conditions.
• The FER performance decreases with each EM iteration, indicating effective joint estimation and detection, though bit errors may not monotonically decrease if convergence is not achieved.
• An error floor appears at high SNR, which can be mitigated by increasing the number of decoder iterations or reducing the coding rate to improve soft decision reliability.
• Using low-rate LDPC codes significantly improves FER performance by enhancing soft decision quality, albeit at the cost of reduced spectral efficiency.
• The proposed EKFS-based algorithm demonstrates robustness across a wide range of phase noise variances, making it suitable for practical high-frequency MIMO systems.