Your browser doesn't support javascript.
loading
Detrending Technique for Denoising in CW Radar.
Lee, In-Seong; Park, Jae-Hyun; Yang, Jong-Ryul.
Afiliação
  • Lee IS; Department of Electronic Engineering, Yeungnam University, Gyeongsan 38541, Gyeongbuk, Korea.
  • Park JH; Department of Electronic Engineering, Yeungnam University, Gyeongsan 38541, Gyeongbuk, Korea.
  • Yang JR; Department of Electronic Engineering, Yeungnam University, Gyeongsan 38541, Gyeongbuk, Korea.
Sensors (Basel) ; 21(19)2021 Sep 24.
Article em En | MEDLINE | ID: mdl-34640697
A detrending technique is proposed for continuous-wave (CW) radar to remove the effects of direct current (DC) offset, including DC drift, which is a very slow noise that appears near DC. DC drift is mainly caused by unwanted vibrations (generated by the radar itself, target objects, or surroundings) or characteristic changes in components in the radar owing to internal heating. It reduces the accuracy of the circle fitting method required for I/Q imbalance calibration and DC offset removal. The proposed technique effectively removes DC drift from the time-domain waveform of the baseband signals obtained for a certain time using polynomial fitting. The accuracy improvement in the circle fitting by the proposed technique using a 5.8 GHz CW radar decreases the error in the displacement measurement and increases the signal-to-noise ratio (SNR) in vital signal detection. The measurement results using a 5.8 GHz radar show that the proposed technique using a fifth-order polynomial fitting decreased the displacement error from 1.34 mm to 0.62 mm on average when the target was at a distance of 1 m. For a subject at a distance of 0.8 m, the measured SNR improved by 7.2 dB for respiration and 6.6 dB for heartbeat.
Assuntos
Palavras-chave

Texto completo: 1 Coleções: 01-internacional Base de dados: MEDLINE Assunto principal: Radar / Processamento de Sinais Assistido por Computador Idioma: En Revista: Sensors (Basel) Ano de publicação: 2021 Tipo de documento: Article País de publicação: Suíça

Texto completo: 1 Coleções: 01-internacional Base de dados: MEDLINE Assunto principal: Radar / Processamento de Sinais Assistido por Computador Idioma: En Revista: Sensors (Basel) Ano de publicação: 2021 Tipo de documento: Article País de publicação: Suíça