Research On Phase Noise Effects And Mitigation In OFDM Systems

In comparison to single-carrier transmission, Orthogonal frequency division multiplexing (OFDM) is quite effective to eliminate inter-symbol interference (ISI) caused by channel multipath fading hostility, while providing high transmission data rate with high spectral efficiency. Moreover, OFDM receiver becomes relatively simple with one-tap channel equalizer and simple hardware realization. Hence, OFDM is very well suited to the future high data rate wireless multimedia communications.Due to the long symbol duration, one drawback of OFDM is its sensitivity to phase noise introduced by a receiver local oscillator (LO). Two phase noise effects are distinguished: common phase error (CPE) and inter-carrier interference (ICI). Motivated by what has been proposed in literatures, CPE and ICI are described by the exact mathematician formula and analyzed using the proposed effective filter method. In order to investigate phase noise effects on OFDM performance, the exact signal to interference plus noise ratio (SINR) expressions are derived in this dissertation for arbitrary phase noise levels, and for both Wiener phase noise and the stationary Gaussian colored noise (two mathematician models for phase noise). The analytical results make it easier to determine OFDM system parameters.The conventional phase noise processing method, called CPE correction (CPEC), deals with only CPE. However, the correction of CPE does not suffice always, especially when high data rates are required. Hence, it is very necessary to estimate and compensate the phase noise ICI.In this dissertation, a new analytical model, called the frequency-domain linearized parametric model for phase noise, is created. This model can provide better approximation than the direct one-order linearized approximation for phase noise. Based on the proposed model, the performance loss caused by phase noise ICI is analyzed and the SINR expression is derived. Subsequently, a novel phase noise mitigation method, called ICI reduction (ICIR), is proposed to correct both CPE and ICI. The ICIR scheme inserts several phase-noise estimation pilots into the OFDM transmitted band, estimates phase noise parameters at the receiver by LS method according to the disturbed pilots, and finally mitigates phase noise effects in the frequency or time domain. It is shown by simulation results that the ICIR algorithm is very effective and outperforms the conventional methods.The phase noise mitigation in OFDM over Rayleigh fading channels is also investigated in this dissertation. Due to phase noise ICI lowering the effective SNR, it is quite necessary to reduce ICI before channel estimation. It is proved that phase...