基于CAZAC序列的OFDM时频同步算法

An OFDM Timing and Frequency Synchronization Algorithm Based on CAZAC Sequence

针对Schmidl和Minn算法存在符号定时和频率同步精度不高,频率同步运算量大的缺点,提出一种新的基于CAZAC序列的OFDM同步算法。算法利用2个相同的CAZAC序列作为一个训练符号,并对第二个序列用CAZAC序列进行加权,在接收端利用训练符号所包含数据的重复性和CAZAC序列的相关性,进行符号定时同步;然后利用CAZAC序列具有的时域移位特性进行整数倍频偏估计。算法在时域完成,因运算量较小。仿真结果表明,新算法克服了Schmidl和Minn算法的缺点,在AWGN和多径信道下,均能精确同步并进行较大范围的频偏估计。特别是在信噪比较低时采用该算法能实现更准确的同步。

The Schmidl and the Minn algorithm have some shortcomings, such as low precision of symbol timing and Frequency Synchronization and large operand of Frequency Synchronization. A novel algorithm based on CAZAC sequence is proposed for OFDM synchronization. The algorithm uses one unique training symbol which has two identical halves and the second haft is weighted by CAZAC sequence. At the receiving end the symbol timing synchronization is accomplished by using the repeatability of the training symbol and the correlation of the CAZAC sequence. Then the integral frequency offset estimate is accomplished by using the time domain shifting characteristic of the CAZAC sequence. This algorithm is achieved in time domain, so the operand is small. Simulation results show that the proposed algorithm has overcome the shortcomings of Schmidl and Minn algorithm, and can synchronize precisely and carry out wide range frequency offset estimate in AWGN and muhipath fading channel. Specially, this algorithm can realize a more accurate synchronization.