一种基于PN训练序列的OFDM符号定时算法

摘要

作为第四代移动通信的核心技术,正交频分复用(OFDM)系统越来越受到大家的关注.而OFDM同步技术则是OFDM系统中的一项关键技术.针对Schmidl&Cox提出的同步算法存在符号定时很难准确测定的缺点,提出了一种新的基于PN训练序列的OFDM符号定时算法.在Schmidl&Cox算法基础上,重新对训练序列的第一个符号进行设计,并对符号前半段按伪随机序列进行重排.从理论分析和仿真结果表明,时间度量曲线会出现单一"脉冲"值,避免Schmidl&Cox算法产生的平台和平缓的峰.与传统符号定时方案相比,算法同步性能更好,符号定时更精确,是一种非常有效实用的方案.%As a key technology for the 4th generation wireless mobile communication system, OFDM has drawn more and more attention and much work has been done for its synchronization methods. The paper proposes an efficient symbol timing synchronization scheme for OFDM system which makes use of PN training sequences to overcome the shortcomings of the synchronization algorithm proposed by Schmidl&Cox. Based on the Schmidl&Cox algorithm, the proposed algorithm redesigns the first symbol of the training sequence and rearranges the front half part of the symbol using PN sequence. Theoretical analysis and simnlation results indicate that the metric curve may generate a single pulse and avoids the presence of the plateau. The proposed scheme is very efficient, and compared with the conventional schemes, it not only has a better performance in frequency synchronization, but also is more accurate in symbol timing.