| 摘要: |
|
| 关键词: 卫星导航系统 多径效应 多径抑制 残差网络 通道注意力 |
| DOI：10.20079/j.issn.1001-893x.231204003 |
|
| 基金项目:湖北省自然科学基金面上项目（2024AFB933） |
|
| A Multipath Detection and Suppression Method Based on Residual Network |
| LI Zhichao,ZHOU Yanling |
| (School of Artificial Intelligence，Hubei University，Wuhan 430062，China) |
| Abstract: |
| The multipath effect in Global Navigation Satellite System(GNSS) is closely related to the environment，and it is difficult to be eliminated by the model，which has become a bottleneck problem to improve the positioning and navigation performance.For the problem that traditional methods are difficult to solve the multipath effect in complex environments，a multipath detection and suppression method based on efficient channel attention deep residual network(ECA-ResNet) is proposed.The model combines residual connection and attention mechanism to extract multipath component features from the relevant images at the receiving end and obtain relevant multipath parameters.The multipath parameters are stacked with the relevant images，and the neural network is used to adaptively adjust the relevant images to achieve multipath signal suppression.The experimental evaluation is carried out on the GNSS generator developed by Ecole Nationale de l Aviation Civile(ENAC).The results show that the proposed ECA-ResNet multipath detection method improves the determination coefficient of the three multipath parameters on the basis of the traditional residual network method by 0.3%，0.74% and 4.24%，respectively.At the same time，the average structural similarity index of the two branch-related images after multipath suppression is increased by 4.34% and 4.76%,respectively. |
| Key words: satellite navigation system multipath effect multipath suppression residual network channel attention |
《电讯技术》编辑部 