Modulation type classification of interference signals in automotive radar systems. Issue 6 (1st June 2019)

• Record Type: Journal Article
• Authors: Kim, Jinwook; Lee, Seongwook; Kim, Seong‐Cheol
• Is Part Of: IET radar, sonar & navigation. Volume 13:Issue 6(2019)
• Page Start: 944
• Page End: 952
• Publication Date: 2019-06-01
• Subjects: radar signal processing -- support vector machines -- low‐pass filters -- road vehicle radar -- radio receivers -- signal classification -- interference (signal) -- mixers (circuits) -- object detection -- interference suppression; interference signals -- mitigating interference effects -- complete interference mitigation algorithm -- different radar modulation signals -- automotive radar systems -- mutual interference; Signal processing -- Periodicals; Radar -- Periodicals; Sonar -- Periodicals; Electronics in navigation -- Periodicals; Navigation -- Periodicals; 621.3848
• DOI: 10.1049/iet-rsn.2018.5521
• Languages: English
• ISSNs: 1751-8784

Abstract

Abstract : As the importance of radar sensors for automotive vehicles has rapidly increased, the mutual interference between vehicle radars has become important. If two or more radars share the same transmitting frequency band, their interference effects are retained after low‐pass filtering. These interference signals can degrade the performance of target estimation, as they increase the noise floor level and create ghost targets, consequently causing false alarms. Several studies on mitigating interference effects have been conducted. However, they only focused on suppression techniques. For a complete interference mitigation algorithm, the identification of the existence of interference signals and classification of their modulation techniques should precede the mitigation steps. Therefore, the authors propose a method for the identification and the classification of interference signals. They modelled five different radar modulation signals as interference signals, and extracted five features from the time‐domain mixer output signals. The training of the classification model and validation of its classification performance is done using the support vector machine. The simulation results show that the existence of interference signals is identified with the accuracy of over 99%. Furthermore, the interference signals are effectively classified into their modulation types with an accuracy of over 96%.