论文部分内容阅读
在正交频分系统提出一种频偏估计的最优训练序列结构,和传统训练序列相比在没有变化总能量的情况下Cramer-Rao lower bound(CRLB)达到最小。最优训练序列结构利用最大似然算法(ML)进行频偏估计,通过计算机仿真可以看出这种频偏估计和传统的频偏估计相比的优点在于相同的算法复杂度情况下频偏估计更精确。
In the orthogonal frequency division system, an optimal training sequence structure of frequency offset estimation is proposed, which achieves the minimum Cramer-Rao lower bound (CRLB) without changing the total energy compared with the traditional training sequence. The optimal training sequence structure uses the maximum likelihood algorithm (ML) to estimate the frequency offset. The computer simulation shows that the advantage of this frequency offset estimation compared with the traditional frequency offset estimation is that the frequency offset estimation under the same algorithm complexity more accurate.