期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2014
卷号：3
期号：11
页码：4060-4065
出版社：Shri Pannalal Research Institute of Technolgy
摘要：Experimental results shows that absolute error is decreases when number of iterations increases for all models among that WL-QLMS has less absolute error compares to other methods. The quaternion least mean square (QLMS) calculation is presented for adaptive shifting of three- and four-dimensional methods, for example, those can be observed in environmental displaying (wind, vector fields). These techniques display complex nonlinear elements furthermore coupling between the measurements, which make their segment knowledge handling by different univariate LMS, bivariate complex LMS (CLMS), or multichannel LMS (MLMS) calculations insufficient. The QLMS represents these issues characteristically, as it is determined straightforwardly in the quaternion space. The dissection demonstrates that QLMS works inherently focused around which is based on "increased" detail, that is, both the covariance and pseudo covariance of the tap data vector are considered.
关键词：Properness; quaternion adaptive filtering; ; quaternion LMS (QLMS); AQLMS (argumented quaternion ; least mean square; widely linear model; widely linear ; quaternion least mean square model
