| 摘要: |
| 本文研究了基于BP神经网络的通信信号调制体制识别技术。提取了反映信号调制体制差异的特征参数，并从提高收敛速度和正确识别概率出发，构建了最佳的神经网络分类器。实验结果表明，该技术拥有较好的通信信号调制体制识别性能。 |
| 关键词: 通信信号 调制识别 特征参数 BP神经网络 组合网络结构 大数判决 |
| DOI：10.3969/j.issn.1001-893X. |
| 投稿时间：2005-05-18修订日期：2005-10-12 |
| 基金项目: |
|
| Modulation Recognition Technology for Communication Signals Based on BP Neural Networks |
|
| () |
| Abstract: |
| The modulation recognition technology for communication signals based on BP neural networks is studied. The character parameters reflecting the difference of signal modulations are extracted. According to convergence rate and right identifiable probability, the best neural networks classification is established. Experiment result shows that these technologies have better performances for automatic modulation recognition of communication signals. |
| Key words: communication signal modulation recognition character parameter BP neural network combined networks structure judgement of the biggest number |
《电讯技术》编辑部 