正交频分复用（OFDM）技术以其高的频谱利用率和良好的抗多径干扰能力，在无线通信领域得到广泛应用，它将是未来无线宽带接入系统的基本技术之一，被看作是3G以后移动通信的主流技术。载频偏移问题是OFDM系统的突出缺点，它使系统信噪比下降，破坏子载波之间的正交性，带来严重的子信道间干扰（ICI），严重影响系统性能。所以载频偏移问题是OFDM技术得到应用的关键之一，对频偏校正算法的研究成为近年来人们研究的热点问题。 本论文围绕在理论和实践上有广泛意义的两大类频偏校正算法：数据辅助即基于导频符号估计算法和非数据辅助即盲估计算法展开如下研究： （1）介绍了OFDM技术的基本原理、优缺点和主要技术。详尽分析了载频偏移对系统性能的影响。 （2）对数据辅助即基于导频符号估计算法的经典算法Sch-Cox算法及其改进算法进行了深入的算法分析和性能比较。提出自己的改进算法—将训练帧序列设计成实伪随机（PN）序列进行频偏校正和定时估计的方法，该改进算法进一步提高了定时和频偏估计的精度，扩大了频偏估计范围。即使在多径信道中也具有很好的定时估计精度和频偏估计范围，可以应用到实际的系统中。 （3）对非数据辅助即盲估计的四种算法：基于循环前缀的最大似然估计算法、两种基于子空间的算法：MUSIC_Like算法和ESPRIT_Like算法、基于循环平稳特性的估计算法进行了详尽的算法描述和性能仿真。 （4）对各种频偏校正算法进行了综合仿真和比较分析。Sch-Cox算法及其改进算法估计精度最高；MUSIC-Like和ESPRIT_Like属于盲估计，其估计精度次之；基于循环平稳特性算法的估计性能和估计精度介于Sch-Cox算法和MUSIC-Like算法之间；基于循环前缀的最大似然估计算法估计精度较差。

Because of its high spectrum utilization and good capability of preventing from muti-path interference, OFDM has been extensively used in the area of radio communication systems it is regarded as the core technology in Beyond 3G mobile communication system. However, carrier frequency offset is the main drawbacks of OFDM system, it can make the SNR lower, cause loss of sub-carrier orthogonality, and thus can introduce inter-channel interference (ICI) and severely degrade the system performance. Carrier frequency offset is the key problem of the use of OFDM. So researchers recently focus on the study of the correcting algorithms of carrier frequency offset. In this paper, the basic principles, the core technologies and the advantages of OFDM are introduced, the effect of carrier frequency offset is proposed. The algorithm based on the aid of pilot symbols and its improved algorithms are proposed, through the simulations, the conclusions are thus addressed that the improved algorithms can outperform the traditional ones in Estimation precision and ranges. Then, Analyses and compares Four Correcting of Carrier Frequency Offset methods—maximum-likelihood based on CP, MUSIC-Like, ESPRIT-Like and Cyclostationarity. Finally, the simulation results employing different algorithms are compared, the performance of Sch-Cox algorithm and its improved algorithm are best, the blind estimation algorithms are better, Cyclostationarity algorithm among Sch-Cox algorithm and blind estimation algorithm, maximum-likelihood based on CP algorithm is less.