基於正交分頻多工的直接轉換接收器中載波頻率飄移與實虛部非協調與通道響應之估測

Abstract

Carrier frequency offset (CFO) and in-phase and quadrature-phase (I/Q) imbalance are two of the common front-end imperfections in low-cost communication devices. It is known that CFO is a crucial problem in orthogonal frequency division multiplexing-based (OFDM-based) systems. It can destroy the orthogonality between subcarriers and cause significant degradation in system performance. Also, the existence of the I/Q imbalance usually reduces the accuracy of CFO estimation. In the first part of the thesis, we study data-aided scheme for joint CFO and I/Q imbalance estimation and also subspace-based blind CFO estimation. Firstly, we propose a new data-aided scheme for the joint estimation of CFO and I/Q imbalance using simple matrix formulation. The proposed algorithms utilize only the periodicity of the training sequence. They do not need to know the channel impulse response and the exact values of the training sequence. Analytic performance evaluation of the proposed algorithm is also carried out. The proposed method compares favorably with the existing methods in terms of both bit-error-rate (BER) performance and complexity. Secondly, we propose a novel subspace-based blind CFO estimation method for cyclic-prefixed OFDM (CP-OFDM) systems. Our method is based on the concept of blind subspace channel estimation algorithm using only a few received CP-OFDM blocks in one of the earlier works. The CFO is obtained by solving the nullspace of the proposed rank-reduced matrix and the CFO estimate is given in analytic form. We also prove that the proposed method can work properly in CP-OFDM systems with and without virtual carriers (VC), which is independent of how VC positions are assigned. Moreover, we do not make the assumption that the modulation symbols are white or constant-modulus. Simulation results show that the proposed method not only has a good performance, it is more robust to various signal constellations and VC assignments than the existing methods. In the last decade, several subspace-based algorithms for blind channel identification were introduced. These subspace-based methods are attractive because highly accurate estimate of the channel can be obtained by using only a few received blocks. However, these subspace-based methods do not work when applying to the zero-padded orthogonal frequency division multiplexing systems (ZP-OFDM) with VC. In the second part of the thesis, we propose two improvements on an earlier subspace-based blind channel estimation method. Firstly, we introduce a simple noise weighting matrix. This weighting matrix is diagonal and it is independent of the channel noise variance and SNR. By doing so, the performance of previous work is significantly enhanced. Secondly, we extend the blind estimation method to ZP-OFDM systems with virtual carriers. The proposed method not only works for ZP-OFDM system with VC, but also exploits the VC information to further enhance the estimation accuracy.