多径信道下OFDM信号和单载波信号的盲识别

Blind Recognition of OFDM Signal and Single-Carrier Signals in Multi-Path Channel

针对非协作通信条件下OFDM(正交频分复用)信号和单载波调制信号的类间识别问题,提出一种在多径信道下基于信号循环自相关的调制识别算法。该算法不需要信号和噪声的先验知识,可以直接对中频信号进行处理,具有较强的实用性。通过对信号循环自相关函数进行分析,仅需进行时域的延时自相关运算和特定延时下的循环自相关运算,就可以从频域提取特征值来实现OFDM信号和单载波信号的盲识别。在估计OFDM信号周期时,提出一种基于极大值抽取的延时估计算法,能在多径衰落信道中不需要先验信息的情况下完成估计。仿真表明,本文提出算法的识别性能优于传统的四阶累积量算法。

In order to solve the inter-class blind recognition problem of 0FDM （orthogonal frequency division multiplexing） signals and digital single-cartier signals for non-cooperative communication applications, a blind identification method is proposed based on cyclic auto-correlation under the condition of multi-path fading channel background. The method can be used to directly process bandpass signals without the prior information about the signals or noise, so it has strong practicability. By analyzing the cyclic autocorrelation function of signals, only an autocorrelation operation in the time domain and a cyclic autocorrelation operation under the specific delay are required, then a new feature is extracted to complete the blind recognition. An estimation algorithm based maximum extraction is proposed to estimate the OFDM signal period, and can work well in the multi-path fading channel without prior information. Simulation results show that the proposed algorithm has better performance than the conditional algorithm based fourth-order cumulants.