| 基于小波域滤波的电子通信信道恶意干扰信号分离方法 |
投稿时间：2021-09-06 修订日期：2021-09-16  点此下载全文 |
| 引用本文： |
| 摘要点击次数: 92 |
| 全文下载次数: 0 |
|
|
| 中文摘要:针对电子通信信号与恶意干扰信号严重重叠的现象，本研究设计了基于小波域滤波的电子通信信道恶意干扰信号分离方法。该方法在建立信道模型的基础上，分析信道特征并划分噪声种类。在明确噪声信号形式后，利用小波域滤波算法合理设置分解层数，对信号做分解重构，再通过惩罚标准选取阈值，对信号做阈值降噪，从而确保信号具有显著特征。然后利用极限学习机建立神经网络学习模型，根据干扰信号模型与信道衰减程度提取干扰信号特征，并将特征样本输入到神经网络中，直到输出分离结果。仿真结果表明：该方法能够有效去除信道噪声、降低通信误码率、均衡信道负载，从而证明了该方法在分离干扰信号方面的有效性。 |
| 中文关键词:小波域滤波 电子通信信道 恶意干扰信号 信号分离 信号去噪 |
| |
| Separation method of malicious interference signals in electronic communication channel based on wavelet domain filtering |
|
|
| Abstract:In view of the serious overlap between the electronic communication signal and the malicious interference signal, this paper designs the separation method of the malicious interference signal in the electronic communication channel based on the wavelet domain filtering. This method analyzes the channel characteristics and divides the noise types on the basis of establishing the channel model. After defining the form of noise signal, the wavelet domain filtering algorithm is used to set the decomposition layer reasonably, decompose and reconstruct the signal, and then the threshold value is selected by the penalty standard to de-noise the signal, so as to ensure that the signal has significant characteristics. Then the learning model of neural network is established by extreme learning machine, and the characteristics of interference signals are extracted according to the interference signal model and channel attenuation degree, and the feature samples are input into the neural network until the results of separation are output. Simulation results show that the proposed method can effectively remove channel noise, reduce communication bit error rate and balance channel load, which proves the effectiveness of the proposed method in the separation of interference signals. |
| keywords:Wavelet domain filtering Electronic communication channel Malicious interference signal Signal separation Signal denoising |
| 查看全文 查看/发表评论 下载pdf阅读器 |
位访问者,今日访问
