TY - JOUR
T1 - An approach for extracting independent micro-Doppler characteristics of multiple targets based on underdetermined blind source separation
AU - Guo, Kunyi
AU - Zhang, Yongli
AU - Sheng, Xinqing
AU - Shen, Ronghui
AU - Jin, Congjun
PY - 2012/8
Y1 - 2012/8
N2 - Nowadays the existing approach to extract various micro Doppler characteristics from backscattered signals by multi-targets usually employ blind source separation technique based on independent component analysis. However, this approach has a great limitation in real application. First, micro Doppler characteristics to be separated must be statistically independent; second, independent component analysis method is generally limited to the conditions of exact determined or over determined equations. Under the radar scene with multiple targets, the number of radar signal receivers is commonly less than the number of targets, as well as the micro Doppler characteristics may not satisfy the independent condition, which makes micro Doppler characteristics extraction more difficult. In order to solve these problems, an approach based on underdetermined blind source separation method is pro-posed in this paper. The approach not only work well for the situation that the number of received radar signals is less than that of targets, but also has fewer limitations in the micro Doppler characteristic to be separated. The simulation presented in this paper validates the proposed extraction approach.
AB - Nowadays the existing approach to extract various micro Doppler characteristics from backscattered signals by multi-targets usually employ blind source separation technique based on independent component analysis. However, this approach has a great limitation in real application. First, micro Doppler characteristics to be separated must be statistically independent; second, independent component analysis method is generally limited to the conditions of exact determined or over determined equations. Under the radar scene with multiple targets, the number of radar signal receivers is commonly less than the number of targets, as well as the micro Doppler characteristics may not satisfy the independent condition, which makes micro Doppler characteristics extraction more difficult. In order to solve these problems, an approach based on underdetermined blind source separation method is pro-posed in this paper. The approach not only work well for the situation that the number of received radar signals is less than that of targets, but also has fewer limitations in the micro Doppler characteristic to be separated. The simulation presented in this paper validates the proposed extraction approach.
KW - Continuous wave radar
KW - Micro-Doppler
KW - Sparseness
KW - Time-frequency representation
KW - Underdetermined blind source separation (UBSS)
UR - http://www.scopus.com/inward/record.url?scp=84867435148&partnerID=8YFLogxK
M3 - Article
AN - SCOPUS:84867435148
SN - 1005-0388
VL - 27
SP - 691-695+759
JO - Dianbo Kexue Xuebao/Chinese Journal of Radio Science
JF - Dianbo Kexue Xuebao/Chinese Journal of Radio Science
IS - 4
ER -