摘要： 针对高斯白噪声信道下通信信号的频率估计问题，提出一种基于支持向量机的频率估计算法。利用支持向量机的稳健性和泛化性将频率估计转化为小样本分类问题，使用较少的导频符号提高频率估计性能。该算法不需要接收数据的统计信息，对信号的初始相位不敏感，且不存在门限效应。仿真结果表明，该算法的频率估计性能在低信噪比下优于最大似然估计算法。

关键词: 支持向量机, 频率估计, 模式识别, 最大似然估计, 门限效应

Abstract: This paper addresses the problem of carrier frequency recovery in additive noise. Frequency estimation algorithm based on Support Vector Machine(SVM) is derived. The estimator can work efficiently without the need of statistics knowledge of the observations, and the estimation performance is insensitive to the carrier phase; it shows a better performance than traditional Maximum Likelihood(ML) estimation algorithm at low Signal to Noise Ratio(SNR), for SVM-FEA has not the threshold effect.

Key words: Support Vector Machine(SVM), frequency estimation, pattern recognition, Maximum Likelihood(ML) estimation, threshold effect

中图分类号: 

• TN91