Carrier Frequency Offset And Channel Estimation Algorithms In Mobile Broadband

With the improvement of people's living standard, the requirement of data transmission speed in mobile communication systems is getting higher and higher.In order to meet the increasing demand, the traditional single-carrier communication systems has been unable to avoid the design of complex equalizer problem. Therefore, multi-carrier technology has been widely adopted in the current high-speed broadband mobile communication systems. After several decades of development, numerous achievements have been made in multi-carrier systems. In this thesis, based on the analysis of the existing CFO and channel estimation algorithms, improved algorithms for the two problems are proposed.Recently, OFDMA has been adopted as a multi-user transmission technique in IEEE 802.16, 3GPP LTE and other standards. However, as the fact that the CFO can destroy the orthogonality among subcarriers, it is essential to estimate CFO and compensate at the receiver before signal detection. Hence, by using multi-layer method, in this thesis, an ML-DBCS multi-user blind CFO estimation algorithm is presented on the basis of DBCS based CFO estimation algorithm to reduce the computational complexity. The CRB is also derived to evaluate the performance. In addition, by analyzing the computational complexity of the proposed algorithm, the scheme of determining the optimal number of layers is provided in the thesis. To make the algorithm more practical, the second-order and third-order approximations of the optimal number are calculated.GFDM is a promising solution for the cellular system of 5G PHY layer due to its flexible configuration of pulse shaping filters and subcarriers, good spectrum property, reduced CP overhead and simple equalization. However, because of the pulse shaping, the inherent interference existing in the GFDM symbols, has a negative impact on pilot-based channel estimation. To eliminate the interference on the pilot symbols, an iterative method combined with CS based channel estimation is used, which can efficiently improve the accuracy of channel estimation. Moreover, the CRB is derived to theoretically verify the effectiveness of the proposed algorithm.Finally, the simulation results show that the proposed ML-DBCS algorithm only has a slight performance loss at low SNR, which is caused by the error propagation introduced from the former layers. However, it can reach a maximum complexity reduction of approximate two orders of magnitude, and it also demonstrated that the second and third approximations of the optimal number of layers also have high accuracy. In this thesis, the characteristics of GFDM are analyzed in detail by conducting simulations under the scenarios of different pulse shaping filters, roll-off factor, detection algorithm and subcarrier configuration over AWGN channel. In addition, from the simulation results of different channel estimation algorithms in GFDM system, it can be seen that the proposed iterative channel estimation algorithm can obtain higher channel estimation accuracy and lower BER in different configurations of roll-off factor and modulation scheme.