几种调制分类方法的原理与仿真试验研究

准对数似然函数比 (qLLR)、相位统计矩和不同阶幅度矩是调制分类的新方法。用qLLR方法构造检测MPSK信号的假设检验及其对数似然函数比 ,并求其数学期望 ,得到qLLR以BPSK/QPSK信号分类。相位统计矩方法利用MPSK信号符号初相位偶阶统计矩为M的单调递增函数的特性 ,并进行M进制识别复合假设检验。不同阶幅度矩进行调制分类的方法利用MPSK/MQAM信号每符号段波形起始点与其它点幅度的乘积统计矩特征进行调制分类。给出了计算机仿真试验评估。

Study on Some Approaches to Modulation Classification and Simulations

The qLLR?based classification, classification using statistical phase moments(SPM) and classification with different order amplitude moments are new methods for modulation classification. In the qLLR a hypothesis test and its logarithmic likelihood function are constructed, and the mathematical expectation of the latter is worked out in order to classify BPSK/QPSK signals. In the SPM method the composite hypothesis test is performed by means of the monotonic characteristic of the even order statistical phase moments of MPSK signals. The method with different order amplitude moments makes use of the product moment of signal magnitude for classification of MPSK/MQAM signals. Through simulation, the classification performance of the methods is assessed.