| 引用本文： |
-
姚元飞,邱吉刚,张小舟,等.基于Sinc函数和径向基神经网络的非线性干扰对消方法[J].电讯技术，2026，66（3）： - . [点击复制]
- YAO Yuanfei,QIU Jigang,ZHANG Xiaozhou,et al.Nonlinear Interference Cancellation Method Based on Sinc Function and Radial Basis Function Neural Network[J].，2026，66（3）： - . [点击复制]
|
|
| 摘要: |
| 针对民航空管密集大功率信道布局引起的非线性邻道共址干扰日益严重的问题，提出了一种基于Sinc函数和径向基神经网络（Radial Basis Function Neural Network，RBFNN）的模数结合干扰对消方法。首先，通过Sinc函数和梯度优化算法，建立非线性模拟对消模型，消除多径影响，并在一定程度上对消共址干扰；然后，通过RBFNN融合邻道信号检波特征和干扰残差特征，建立非线性数字对消模型，进一步对消残留干扰。实验表明，该方法可以在有效保留接收信号中有用信号的前提下快速消除非线性共址干扰。在干信比为60 dB的条件下，干扰对消比可以达到99.7 dB，相对现有方法至少提升了24 dB；有用信号损失度仅为11.9 dB，相对现有方法至少减小了18 dB；干扰对消时间仅为70 μs，相对现有方法至少缩短了35 μs。 |
| 关键词: 空中交通管理 非线性干扰 共址干扰 干扰对消 径向基神经网络（RBFNN） |
| DOI：10.20079/j.issn.1001-893x.240829001 |
|
| 基金项目:四川省科技成果转移转化示范项目（2024ZHCG0046） |
|
| Nonlinear Interference Cancellation Method Based on Sinc Function and Radial Basis Function Neural Network |
| YAO Yuanfei,QIU Jigang,ZHANG Xiaozhou,CAI Fangkai,WU Jianguang |
| (1.School of Electrical and Electronic Engineering，Chengdu Technological University，Chengdu 611730，China;2.Chengdu Spaceon Technology Co.，Ltd.，Chengdu 611731，China) |
| Abstract: |
| In view of the increasingly serious problem of nonlinear adjacent channel co-location interference caused by dense high-power channel layout in civil aviation air traffic control，a combined analog-digital interference cancellation method based on cardinal sine(Sinc) function and radial basis function neural network(RBFNN) is proposed.Firstly，through Sinc function and gradient optimization algorithm，a nonlinear simulation cancellation model is established to eliminate multipath influence and to a certain extent eliminate co-location interference.Then，by fusing the detection features of adjacent channel signals and interference residual features through RBFNN，a nonlinear digital cancellation model is established to further cancel residual interference.Experiments have shown that this method can quickly eliminate nonlinear co-location interference while effectively preserving the useful signal in the received signal. Under the condition of 60 dB interference-to-signal ratio，the interference cancellation ratio can reach 99.7 dB，which is at least 24 dB higher than that of the existing methods；the loss degree of useful signal is only 11.9 dB，which is at least 18 dB lower than that of the existing method；the interference cancellation time is only 70 μs，which is at least 35 μs shorter than that of existing methods. |
| Key words: air traffic control nonlinear interference co-location interference interference cancellation radial basis function neural network(RBFNN) |
《电讯技术》编辑部 