Estimation And Mitigation Of The Phase Noise And Carrier Frequency Offset In Wireless Communication System

The phase noise(PHN) and carrier frequency offset(CFO) which are caused by the jitter of oscillators and the Doppler effect induced by the movement between transmitters and receivers are two of the most important factors that disturb the performance of the receivers in wireless communication. In this thesis these two parameters were investigated and several algorithm were proposed for different kinds of communication systems. For the OFDM system and single carrier with cyclic prefix (SCCP) system the algorithm that estimates the PHN by exporting DCT and VB method was proposed while for low earth orbit (LEO) satellite receivers we proposed new particle filter based algorithms that improves the performance of two-step CFO estimator by two ways. The main contributions of this thesis are as follows:1. The effect of the PHN for OFDM system was investigated as well as the exiting estimation method. Based on the method which exploits DCT and LS approach, we derived three new methods utilizing DCT and VB approach with the first one, however, being equivalent to the LS approach. The other two methods outperform the LS approach since they can suppress the effect of AWGN while estimating the PHN and lead to more accurate estimating result and better system performance. This conclusion was proved by computer simulations. Another conclusion was that the second VB based method can achieve best tradeoff between performance and complexity. 2. A DCT based PHN estimation method was proposed for SCCP system. based on the feature of PHN, we proposed to approximate the PHN as a sum of discrete time domain components which linearized the effect of the PHN. Then LS estimation was exploited to estimate the DCT transmitted symbols which was equally to estimate the PHN in time domain. Then, the VB method was proposed to jointly estimate the transmitted symbols and the variance of AWGN by iterative mean considering that the LS approach failed to mitigate the effect of AWGN while estimating PHN. The proposed methods were compared by computer simulations.3. We investigated the frequency offset tracking problem for MPSK DS-CDMA system. A non-linear non-Gaussian model was firstly formulated by considering the MPSK DS-CDMA signal structure and the frequency-offset-induced power degradation. Then a particle filtering based frequency offset tracking method was proposed to solve the model. The simulation results show the proposed method can achieve better performance than the existing gradient descent method in tracking accuracy in low SNR scenario while in high SNR both of the two method performed similarly.4. A new automatic frequency control (AFC) scheme was proposed, which could be used for the receiver of LEO satellite communication system in continuous transmitting scenario. By employing the time varying characteristic of particle filter technique, the new scheme combined the preamble based estimating step and data based estimating step to provide initial probability density recursively. It was proved by computer simulations with the criteria of Root-Mean Square (RMS) frequency estimating performance and the bit error rate performance that the proposed AFC scheme could provide better performance than the two-step scheme.