【作者】 赵静；

【导师】 肖丽萍； 张世红；

【作者基本信息】 燕山大学 ， 电子与通信工程， 2014， 硕士

【摘要】 电力线通信(PLC)技术是利用已有配电网进行数据及多媒体信号传输的一种特殊通信方式。正交频分复用(OFDM)是电力线通信的重要调制技术，但其主要缺点之一是对载波频率偏移非常敏感，故载波同步是OFDM系统的关键技术之一。载波频率偏移经子载波间隔归一化后可分为整数频偏(IFO)和小数频偏(FFO)，本文分别对其估计算法进行研究。首先，对OFDM系统整数频偏估计算法进行研究，对原Chu序列互相关整数频偏估计算法进行改进，提出了基于最大后验概率准则的整数频偏估计算法。该算法在原算法基础上考虑训练序列周围的数据符号和噪声影响，增加了一个修正项，可以进一步提高整数频偏估计的准确性。其次，对OFDM系统小数频偏估计算法进行研究，针对基于普通傅里叶变换的传统OFDM系统最大似然(ML)准则载波频偏估计算法在频率选择性衰落信道下性能急剧下降的问题，提出了AWGN信道下基于分数傅里叶变换(FrFT)的OFDM系统ML准则小数频偏估计算法。该算法采用chirp基代替正弦基作为OFDM系统的子载波，可以在不增加计算复杂度的前提下，更有效地对抗频率选择性衰落信道中由小数载波频偏带来的干扰。最后，将FrFT应用于传统OFDM系统的频率选择性衰落信道ML准则载波频偏估计算法，给出了一种频率选择性衰落信道下基于FrFT的小数频偏估计算法。该算法充分利用频率选择性衰落信道的多径时延扩展信息，延长了对循环前缀作相关运算的区间，与AWGN信道下基于FrFT的OFDM系统ML准则小数频偏估计算法相比，能够实现对频率选择性衰落信道下小数载波频偏的更准确估计。上述三种算法的估计性能均在电力线信道下进行了仿真验证。更多还原

【Abstract】 Power line communication (PLC) is a special communication using the existingpower line distribution network for data and multimedia signals transmission. Orthogonalfrequency division multiplexing (OFDM) technology is an important modulationtechnology of PLC, but one of the main shortcomings of an OFDM system is its highsensitivity to the carrier frequency offset (CFO). Therefore, CFO synchronization is oneof the key technologies of an OFDM system. CFO normalized by the subcarrier intervalcan be divided into integral frequency offset (IFO) and fractional frequency offset (FFO).The paper will research the IFO estimation algorithms and FFO estimation algorithmsseparately.Firstly, the paper researches the IFO estimation algorithms, and proposes amaximum-a-posteriori (MAP) criterion IFO estimation algorithm which is an improvedmethod on the basis of the original cross-correlation IFO estimation algorithm with Chusequence. The proposed algorithm takes the data symbols around training sequence andthe noise into consideration on the basis of the original algorithm, and a correction termis added in the algorithm which can improve the IFO estimation accuracy.Secondly, the paper researches the FFO estimation algorithms. Aimming at theperformance’s sharp decline in frequency selective fading channels of themaximum-likelihood (ML) criterion FFO estimation algorithm of the Fourier Transformbased traditional OFDM systems, the paper puts forward the Fractional FourierTransform (FrFT) based OFDM systems ML criterion FFO estimation algorithm underthe AWGN channels. The algorithm adopts chirp basis functions replacing the sinusoidalbasis functions as the OFDM system subcarriers, and can be against the interferencecaused by FFO more effectively than the traditional OFDM algorithm under frequencyselective fading channels on the premise of no computational complexity increasing.Finally, applying the FrFT to the traditional OFDM systems ML criterion CFOestimation algorithm over frequency selective fading channels, the paper give one FrFTbased FFO estimation algorithm of frequency selective fading channels. The algorithm makes full use of the multipath delay spread information of the frequency selectivefading channels, extending the range for the cyclic prefix correlates. And comparied withthe FrFT based OFDM systems ML criterion FFO estimation algorithm in AWGNchannels, this algorithm can estimate the FFO more accurately under frequency selectivefading channels.The estimation performance of the three algorithms is verified through simulationsunder power line channels.更多还原