基于自适应特征模态分解的RFID欠定源数估计方法

打开文本图片集

关键词：盲源分离；同频干扰；欠定；特征模态分解；奇异值分解；Newton-Raphson优化算法；源数目估计中图分类号：TN911.7-34 文献标识码：A 文章编号：1004-373X（2026）07-0103-08

RFID underdetermined source number estimation algorithm based on adaptive feature mode decomposition

WANG Ling'，LUO Zhongqiangl，

（1.ScholofAutomatioandInformationEnineering，SianUnversityofSienceandEngining，Yibin64oCin; 2.IteligentPetidtrolKboatoofaProicanUitfcdginga

Abstract：RFID （radio frequencyidentificationdevice）tagsignalsare prone toco-frequency interference in thesame frequencyband.Theblindsourceseparation（BS）technologyisaneffectivewaytoreduce theinterferenceaforementioned.One ofthenecesaryconditionsfortheimplementationofBSStechnologyistoestimatethenumberofsourcesignalsacurately.In viewof this，asourcenumberestimationmethodbasedonimprovedsingularvaluedecomposition（SVD）withadaptivefeature modedecomposition （FMD）isproposed to make thesourcenumber estimation ofco-frequencytag signalsacurateunder underdeterminedconditions.ThenumberofmodalitiesandfiltersizearecombinedandoptimizedbytheNewton-Raphson optimizationalgorithm，andthentheobservedsignalsaresubjectedtoFMDbytheoptimalcombinationparameters.Themultiple groupsof modalcomponentsarecombined toobtainanew modalfunction，thecovariance matrixof whichissubjectedtoSVD， bywhich theestimatedvalueofthenumberof sources isobtained.Simulationshowsthattheproposed methodcanimprove the estimation accuracy of the number of source signals and has good adaptability and stability.

Keywords：BSS；co-frequency interference;underdetermination；FMD；SVD;Newton-Raphsonoptimizationalgorithm： source number estimation

0 引言

增多，RFID系统中多个标签共用同一频段，在信号传输过程中容易造成同频干扰，难以直接分离。（剩余10551字）