KMS of Xinjiang Astronomical Observatory, CAS
| 基于模板的RFI特征提取与识别方法研究 | |
| Alternative Title | Research on RFI Feature Extraction and Recognition Method Based on Template |
张亚州
| |
| Subtype | 硕士 |
| Thesis Advisor | 张海龙 |
| 2022-05-25 | |
| Degree Grantor | 中国科学院大学 |
| Place of Conferral | 北京 |
| Degree Name | 理学硕士 |
| Degree Discipline | 天文技术与方法 |
| Keyword | RFI，特征提取，特征识别，模板，RFI抑制技术 |
| Abstract | 射频干扰（RFI）作为非观测目标信号，严重影响天文数据质量，如何快速准确识别并抑制射频干扰现已成为射电天文数据处理过程中急需解决的问题。 本文根据实际观测过程中遇到的RFI问题，分析了RFI主动预防策略和滤波方法。针对RFI信号提出了一种基于信号峰值的特征提取方法，利用该方法生成了RFI信号的特征模板，并基于动态时间规整算法和打分策略设计了特征识别算法，实现了RFI信号识别以及分类。 利用ARRL网站[1]提供的RFI数据完成了特征模板的交叉验证对比，测试结果显示RFI特征模板与同类别RFI相似度高，表明该算法能够对RFI进行识别和分类。利用Parkes脉冲星观测基带数据，对实现的算法进行了测试，为验证算法的有效性，在基带数据频域通道中加入ARRL提供的特定RFI信号，利用RFI特征识别算法遍历各频率通道，计算各通道数据与RFI模板的相似度值，并根据阈值判定相应信号类别。实验结果表明本文提出的特征提取方法能够有效提取特定RFI的特征信息，特征识别算法能够根据已有特征模板对数据中混入的RFI进行识别。 主要完成了以下几个方面的工作： （1）研究了RFI主动预防策略和滤波方法，并系统分析了天文观测过程中相应策略或方法的适用范围和优缺点。 （2）设计并实现了基于峰值的RFI特征提取算法。提取相应信号单周期内的多个峰值信息，经平滑处理后作为该信号的特征模板。 （3）设计并实现了基于模板的RFI特征识别算法。利用RFI相似度算法计算未知信号与RFI特征模板相似度大小，通过与预设阈值进行比较，判断该信号是否为RFI。 （4）在真实脉冲星基带数据中混入特定RFI信号，利用特征识别算法有效标记了基带数据中混入的RFI，验证了本文提出算法的正确性。 论文主要创新点如下： （1）提出了基于峰值的RFI特征提取算法，通过算法能够有效提取特定RFI信息并存储为模板，为RFI模板库建立奠定了基础。 （2）提出了基于特征模板的RFI识别算法，根据已有的特征模板实现了ARRL数据和天文基带数据中的RFI信号标记，在数据处理中可通过调整离散值权重，使其具有更高的识别效率。 |
| Other Abstract | As a non-observation target signal, radio frequency interference (RFI) seriously affects the quality of astronomical data. How to identify and mitigate RFI has now become an urgent problem to be solved in the process of radio astronomy data quickly and accurately. According to the RFI problems encountered in the actual observation process, this paper analyzed the RFI active prevention strategies and filtering methods. A feature extraction method based on signal peak is proposed for RFI signal. The feature template of RFI signal is generated by this method, and the feature recognition algorithm is designed based on dynamic time warping algorithm and scoring strategy to realize RFI signal recognition and classification. The cross-validation comparison of the feature templates was completed using the RFI data provided by the ARRL website, and the test results showed that the RFI feature templates have high similarity to the same category of RFI, indicating that the algorithm can identify and classify RFI. To further verify the effectiveness of the algorithm, the specific RFI signal provided by the ARRL official website is added to the frequency domain channel of the baseband data. The RFI feature recognition algorithm is used to traverse each frequency channel, calculate the similarity value between each channel and RFI template, and determine the corresponding signal category according to the threshold. The experimental results showed that the feature extraction method proposed in this paper can effectively extract the feature information of specific RFI, and the feature recognition algorithm can recognize the RFI mixed in the data according to the existing feature template. Mainly completed the following aspects: (1) The RFI active prevention strategies and filtering methods are studied, and the applicable scope, advantages and disadvantages of the corresponding strategies or methods in the astronomical observation process are systematically analyzed. (2) A peak-based RFI feature extraction algorithm is designed and implemented. Multiple peak information in a single cycle of the corresponding signal is extracted and smoothed as the feature template of the signal. (3) A template-based RFI feature recognition algorithm is designed and implemented. The similarity between unknown signal and RFI feature template is calculated by RFI similarity calculation algorithm. By comparing with the preset threshold, the algorithm can judge whether the signal is RFI or not. (4) The specific RFI signal is mixed into the real pulsar baseband data, and the mixed RFI is effectively marked by the feature recognition algorithm, which verifies the correctness of the algorithms proposed in this paper. The main innovations: (1) A peak-based RFI feature extraction algorithm is proposed, which can effectively extract specific RFI information and store it as a template, which lays a foundation for the establishment of RFI template database. (2) A feature template-based RFI identification algorithm is proposed. According to the existing feature templates, the RFI signal recognition in with the ARRL data and astronomical baseband data is realized. In the data processing, the weight of discrete value can be adjusted to make the algorithm have higher recognition efficiency. |
| Pages | 57 |
| Language | 中文 |
| Document Type | 学位论文 |
| Identifier | http://ir.xao.ac.cn/handle/45760611-7/5164 |
| Collection | 研究生学位论文 计算机技术应用研究室 |
| Affiliation | 中国科学院新疆天文台 |
| First Author Affilication | Xinjiang Astronomical Observatory, Chinese Academy of Sciences |
| Recommended Citation GB/T 7714 | 张亚州. 基于模板的RFI特征提取与识别方法研究[D]. 北京. 中国科学院大学,2022. |
| Files in This Item: | ||||||
| File Name/Size | DocType | Version | Access | License | ||
| 基于模板的RFI特征提取与识别方法研究_（3149KB） | 学位论文 | 暂不开放 | CC BY-NC-SA | Application Full Text | ||
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment