Studies On Turbo-Principle-Based Key Techniques Of OFDM And UWB Systems

In order to satisfy people's increasing requirements on communication services, future high-speed wireless communication systems are required to adopt advanced signal processing technologies to improve the spectral efficiency and data transmission reliability. Recently, OFDM and UWB attract much attention as the two key technologies for future wireless communications. Turbo principle, introduced from the Turbo-code, has been widely applied in coding and signal processing areas. This dissertation focuses on the research of applications of "Turbo Principle" in OFDM and UWB systems with emphases on the iterative algorithm for OFDM phase noise mitigation, LDPC coded UWB systems and iterative multiuser detection for UWB systems. The main contents and contributions of this dissertation are as follows:Firstly, the turbo decoding is studied. Detailed theoretical analysis on turbo process is carried out to illustrate the essence of the turbo principle.Secondly, iterative OFDM phase noise mitigation algorithm is studied. The detailed analysis of the phase noise influence on OFDM systems is presented. Accordingly, two iterative phase noise mitigation algorithms based on Turbo Coded OFDM system and IEEE 802.11 PHY standard are separately proposed. Algorithm proposed based on the Turbo Coded OFDM makes full use of the soft output of Turbo decoder to reconstruct the OFDM symbols, which are then used to estimates the phase noise by LMMSE method. Simulation result shows, the proposed algorithm can eliminate the error floor caused by ICI through few times of iteration. Algorithm based on the IEEE 802.11a PHY is similar to the algorithm proposed for the Turbo Coded OFDM system, but its implementation complexity is greatly reduced and it is applicable both to the Wiener phase noise and PLL phase noise. It makes full use of the four pilot subcarriers to estimate the common phase noise, and hence no additional pilot information is required. Certain mathematical method is used to simplify the matrix computation in the algorithm. Simulation result shows that, the proposed algorithm can also eliminate the error floor effectively through few times of iteration.Thirdly, LDPC coded UWB system is studied. Signal and channel model of UWB system as well as the LDPC code are studied. System model of the TH-BPSK-UWB, TH-PPM-UWB and DS-BPSK-UWB are briefly described and their performance comparison is made through computer simulations. An internally adaptive LDPC Coded UWB-IR scheme is proposed to improve the performance of the traditional UWB-IR system with repetition code, by choosing the code rate of LDPC code and the number of repetition code adaptively. Simulation results under both the AWGN and UWB multipath channel show that, the proposed system outperforms the traditional UWB-IR system significantly and even performs better than the Turbo Coded UWB-IR system at high code rate.Finally, iterative multiuser detection for LDPC coded UWB systems is studied. Based on the UWB discrete time model, various multiuser detection algorithms in UWB systems, i.e., the conventional matched filter multiuser detection, optimum multiuser detection and linear multiuser detection are studied, and their performance is compared by computer simulations. Furthermore, two low complexity iterative multiuser detection schemes for LDPC coded UWB system are proposed. Performance and complexity of the proposed receiver structures are analyzed. Simulation results show that the two proposed systems can provide performance closed to a single-user system with low complexity.