基于理想解的模拟调制识别分析

利用排序待判决信号特征逼近各候选调制方式理想解的程度来实现调制方式识别,该方法不需要确定判决门限,运算量较小。以基本的模拟调制方式为对象,深入分析其识别性能,考察各权值系数对识别概率的影响,可以发现:一般各调制方式均只会被误判成其他某一种调制方式,且不同权值系数对判决结果的影响程度相差较大。因此,可以针对调制方式误判类型寻求新的特征参数,探求最优权值系数来进一步提升所提方法的识别性能。

Classification Analysis of Analog Modulation Based on Ideal Solution

Using the order preference for the degree to which the characteristics of the signal to be determined approximat.ing each candidate modulation mode, the classification method of modulation mode is realized based on the ideal solution.This method has low computation cost, and no decision threshold needs to be determined. Taking the basic analog modula.tion modes as the object, the classification performance was analyzed in depth, and the influence of each weight coefficienton the classification probability was investigated. It could be found that each modulation mode was only misjudged as therespective one of the other modulation modes, and the influence of different weights on the decision results was quite differ.ent. Therefore, we can seek new characteristic parameters for misjudged modulation, seek the optimal weight coefficientsto further improve the classification performance.