分数阶域噪声预测判决反馈均衡方法的研究

针对单载波频域均衡中噪声预测判决反馈均衡结构,介绍了一种基于独特字块结构的分数阶域噪声预测均衡方法,即将接收端的时域信号转换到分数阶域实现噪声预测判决反馈均衡。分数阶域的判决反馈均衡在保持前馈与反馈部分相互独立的同时,还利用了分数阶域信号的灵活性使均衡效果更好;与此同时,利用独特字的已知特性,在接收端将独特字和有用数据分开,再进行分数阶域均衡处理,恢复出原始数据。仿真结果表明,基于独特字的分数阶域噪声预测判决反馈方法相比传统基于循环前缀的均衡方法性能有很大提升。

Prediction decision feedback equalization method based on a fractional order domain noise

Aiming at noise prediction decision feedback equalization structure of single carrier frequency domain e-qualization, this paper introduces a fractional order domain noise prediction equalization method based on the unique word block structure, which means the received time domain signals were converted to fractional order domain to real-ize noise prediction decision feedback equalization. Fractional order domain decision feedback equalization not only keeps the property of mutual independence of feedforward part and feedback part, but also makes use of the flexibility of fractional order domain signal to improve equalization effect;at the same time, by making use of the known charac-teristics of the unique word, useful data and unique word can be separated at the receiving end, and then fractional order domain equalization processing is conducted to restore the original data. The simulation results show that com-pared with the traditional method based on circular prefix, this method of fractional order domain noise prediction de-cision feedback, which is based on the unique characters, has great improvement in equalization performance.