search

bitscott1102 / Automatic-modulation-classification-by-decision-tree-based-on-features-extraction

8 stars 2 forks source link

Please tell me the reference paper #1

Closed jiaoxiang1996 closed 4 years ago

jiaoxiang1996 commented 4 years ago

Please

bitscott1102 commented 4 years ago

Hi there, the papers that I referenced are:

  1. Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification, Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun

  2. 张鸣. 软件无线电接收机中信号调制样式识别算法的研究[D]. 西安科技大学, 2005.

The thresholds are listed as following. However, it depends on the single(the range of amplitude...), what kind of noise you add, etc. In my case, I just plot the pictures and set a clear threshold. In some cases, I want the separation to be clear in the signal with lower SNR, while some may be higher SNR. So the thresholds vary cases to cases.

$$ t\left(m{11}\right)=0.301 ; \t\left(m{12}\right)=0.2345 ; \t\left(m{2}\right)=0.306 ; \t\left(m f{1}\right)=0.21015 ; \t\left(mf{2}\right)=0.4226 ; \ t\left(m p{1}\right) =0.43505 $$