Offline signal processing tools for electromagnetic compatibility applications

• Ano de defesa: 2020
• Autor(a) principal: Artur Nogueira de São José
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Minas Gerais; Brasil; ENG - DEPARTAMENTO DE ENGENHARIA ELÉTRICA; Programa de Pós-Graduação em Engenharia Elétrica; UFMG
• Palavras-chave em Português: Electromagnetic compatibility; Adaptive filtering; Time gating; Particle swarm optimization; Intentional electromagnetic interference; Antenna calibration; Engenharia elétrica; Compatibilidade eletromagnética; Interferência eletromagnética
• Link de acesso: http://hdl.handle.net/1843/34478; https://orcid.org/0000-0001-6985-394X

The electromagnetic compatibility field has gained more and more attention from both academia and industry in the past few years. Following its basic design, testing and mitigation procedures is essential to ensure the correct operation of products and complex infra-structures. In this work, we deal with scenarios where these classical procedures are not sufficient to avoid interference problems or are not accessible due to high implementation costs. To explore this subject, we propose two case studies: antenna calibration in non-ideal test sites and security of railway communications under jamming attacks. To deal with these issues, we developed a new technique based on principles of design and analysis of experiments, called DAE-1, and two new signal processing tools, called ASP-1 and ASP-2. The first two (DAE-1 and ASP-1) were applied to the antenna calibration problem, while the last one (ASP-2) was applied to the intentional electromagnetic interference problem. We designed these tools to be easily handled by an electromagnetic compatibility professional without a signal processing background. In particular, ASP-1 has two routines that automatically set the adaptive filters and time gating parameters. The DAE-1 and ASP-1 techniques were validated after an extensive process involving experimental measurements in four different test sites, three different antennas and a benchmark analysis with two state-of-the-art approaches. In our analyses, we compared three antenna gain curves with anechoic chamber measurements. A performance metric based on the Pearson correlation coefficient indicates performance gains that range from 27% to 30% for DAE-1 and from 21% to 25% for ASP-1. Although with smaller gains, ASP-1 significantly reduces the total test time when compared to DAE-1. Then, we artificially increased the interference levels and evaluated ASP-1 under two different configurations, one of them being inspired by a state-of-the-art approach. The performance gains ranged from 15% to 25%. Finally, we used a computational environment to validate the ASP-2 technique. To do so, we generated a simplified version of the global system for mobile communications-railway (GSM-R) signal with and without jamming contributions. During this stage, we noticed an average bit error rate reduction for signal-to-jamming ratios (SJR) ranging from -2.77 dB to 1.16 dB thanks to the ASP-2 algorithm. These reductions are approximately 10 % when the SJR is -2.77 dB. The threshold SJR level (above which the GSM-R communication can be lost) changed from approximately -0.84 dB to -1.95 dB, expanding thus the operating range of the GSM-R system. This suggests that the adaptive line enhancer is a promising tool to deal with jamming attacks.