Semi-blind CFO estimation and ICA based equalization for wireless communication systems

摘要

In this thesis, a number of semi-blind structures are proposed for Orthogonal Frequency Division Multiplexing (OFDM) based wireless communication systems, with Carrier Frequency Offset (CFO) estimation and Independent Component Analysis (ICA) based equalization. In the first contribution, a semi-blind non-redundant single-user Multiple-Input Multiple-Output (MIMO) OFDM system is proposed, with a precoding aided CFO estimation approach and an ICA based equalization structure. A number of reference data sequences are carefully designed and selected from a pool of orthogonal sequences, killing two birds with one stone. On the one hand, the precoding based CFO estimation is performed by minimizing the sum cross-correlations between the CFO compensated signals and the rest of the orthogonal sequences in the pool. On the other hand, the same reference data sequences enable the elimination of permutation and quadrant ambiguities in the ICA equalized signals. Simulation results show that the proposed semi-blind MIMO OFDM system can achieve a Bit Error Rate (BER) performance close to the ideal case with perfect Channel State Information (CSI) and no CFO. In the second contribution, a low-complexity semi-blind structure, with a multi-CFO estimation method and an ICA based equalization scheme, is proposed for multiuser Coordinated Multi-Point (CoMP) OFDM systems. A short pilot is carefully designed offline for each user and has a two-fold advantage. On the one hand, using the pilot structure, a complex multi-dimensional search for multiple CFOs is divided into a number of low-complexity mono-dimensional searches. On the other hand, the cross-correlation between the transmitted and received pilots is explored to allow the simultaneous elimination of permutation and quadrant ambiguities in the ICA equalized signals. Simulation results show that the proposed semi-blind CoMP OFDM system can provide a BER performance close to the ideal case with perfect CSI and no CFO. In the third contribution, a semi-blind structure is proposed for Carrier Aggregation (CA) based CoMP Orthogonal Frequency Division Multiple Access (OFDMA) systems, with an ICA based joint Inter-Carrier Interference (ICI) mitigation and equalization scheme. The CFO-induced ICI is mitigated implicitly via ICA based equalization, without introducing feedback overhead for CFO correction. The permutation and quadrant ambiguities in the ICA equalized signals can be eliminated by a small number of pilots. Simulation results show that with a low training overhead, the proposed semi-blind equalization scheme can provide a BER performance close to the ideal case with perfect CSI and no CFO.