Research On Capacity And Channel Estimation Of OFDM Communication System

The basic goal of wireless communication development is constant pursuit of the error rate as small as possible (reliability), the channel capacity as large as possible (validity). OFDM (Orthogonal Frequency Division Multiplexing) has higher spectrum efficiency, better anti-interference ability and bandwidth expansibility. It is the most competitive key technology in next generation mobile communication system. This thesis centers on the OFDM technology, try to save system overhead, optimize the performance of system and improve the capacity of system. It focuses on the improvement of channel estimation and system capacity.In the first chapter, the progress of the mobile communications is reviewed. The current progress of capacity and channel estimation of OFDM communication system is introduced in order to meet the following elaboration.It discusses various propagation characteristics of mobile radio channel in detail by analyzing the characteristics of wireless communication fading channel. Small-scale fading of mobile radio channel is the focus of discussion. Effect of Doppler frequency shift on mobile channel is also presented. The performance of OFDM and MIMO OFDM systems in broadband wireless communications are simulated, and the effects varying the system parameters such as the guard interval and the number of carriers are testified.An improved WF algorithm based on specific QoS was proposed. It is more consistent with practical application requirements, and the increase of complexity is not obvious.Many conditions need to be satisfied to achieve the optimal value of WF algorithm, such as adaptive channel coding, adaptive modulation and appropriate power control. In addition, the modulation orders are required to be changed consecutively, and the power allocation which can't be quantified is also required to be consecutive, which are hard to achieve in practical application. But the WF algorithm based on specific QoS can quantify power allocated to each subchannel, and discretize modulation orders. The remaining power after quantification is allocated to next subchannel. The algorithm is more close to ideal situation of WF, and the system can keep certain Qos level.In order to get accurate channel estimation under bad channel environments, an improvement of the OFDM system semi-blind channel estimation algorithm based on Maximum Likelihood Parameter Estimation (MLE) is presented. Applied non-uniform interval pilot arrangement, it proposes a semi-blind channel estimation algorithm of OFDM system based on linear minimum mean-squared error (LMMSE). The impulse response calculated by LMMSE algorithm is regarded as the initial value, and the accurate channel estimation is obtained by several iterative steps. This thesis also proposes the semi-blind channel estimation algorithm based on singular value decomposition (SVD) to simplify the complexity of algorithm. The algorithms proposed can realize accurate channel estimation with less complexity, the advantage and validity are proved by simulation.The thesis discusses the improvements of the two adaptive blind channel estimation algorithms of OFDM system, LMS-Bussgang algorithm and ZF-Bussgang algorithm, which have low complexity. On one hand, it adopts the low complexity Bussgang adaptive algorithm, completes blind channel estimation of OFDM sub-channel according to the criterion of Mean Square Error Criterion and Peak Distortion Criterion. On the other, it uses differential detection, which accelerates the convergence speed and overcomes the error transmission problem brought by Bussgang algorithm.The optimization algorithm of star-QAM modulation constellation is proposed. With criterion of minimized average energy and criterion of maximize minimum Euclidean distance as the basic policies, it presents the optimazition mechanism of star-QAM constellation with fading channel and using Quasi-Coherent demodulation. The algorithm takes advantages of anti-fading capabilities in both OFDM and star-QAM modulation and demodulation and provide valuable information for further studying the potential capability in higher-order modulation-demodulation anti-fading technologies and theories.