Analysis of Pilot-Aided Channel Estimation in OFDM System under Phase Noise

Abstract

One of the major problems for accurate channel estimation in OFDM system is a phase noise. The distortion caused by phase noise is characterized by a common phase error (CPE) term and an inter-carrier interference (ICI) term. Current phase noise suppression algorithms obtain channel state information (CSI) by separating CPE and ICI from channel impulsive response (CIR). However, the received signal is determined by not only CPE and ICI but also CIR. Since all of them are unknown and the number of CPE, ICI, and CIR exceeds the length of observation vector, channel estimation in phase noise environment is an underdetermined problem. It is very difficult to obtain CSI through separating the CPE and ICI from CIR. This paper proposes two novel channel estimators for phase noise based on Maximum Likelihood Estimator and Minimum Mean Square Error Estimator. Since the proposed methods treat the CIR and CPE as a whole unknown parameter and regard the ICI as additive noise, they reduce the number of the unknown parameters and make the problem solvable. The corresponding deterministic and stochastic Cramer-Rao lower bounds (CRLBs) are also derived for phase noise situation which have not been studied in the literature.