[1]陆俊,柯炜,金杰,等.基于KPCA的无线层析成像定位方法[J].南京师范大学学报(自然科学版),2020,43(01):31-39.[doi:10.3969/j.issn.1001-4616.2020.01.006]
　LuJun,KeWei,JinJie,et al.WirelessTomographyPositioningMethodBasedonKPCA[J].JournalofNanjingNormalUniversity(NaturalScienceEdition),2020,43(01):31-39.[doi:10.3969/j.issn.1001-4616.2020.01.006]
点击复制

基于KPCA的无线层析成像定位方法()
分享到：

《南京师范大学学报》（自然科学版）[ISSN:1001-4616/CN:32-1239/N]

卷:
第43卷
期数:
2020年01期
页码:
31-39
栏目:
·物理学·
出版日期:
2020-03-15

文章信息/Info

Title:
WirelessTomographyPositioningMethodBasedonKPCA
文章编号:
1001-4616(2020)01-0031-09
作者:
陆俊1; 柯炜1; 2; 金杰1; 陈梦玲1; 王彦力1; 左浩然1
(1.南京师范大学江苏省光电子技术重点实验室,物理科学与技术学院,江苏南京210023)(2.江苏省地理信息资源开发与应用协同创新中心,江苏南京210023)
Author(s):
LuJun1; KeWei1; 2; JinJie1; ChenMengling1; WangYanli1; ZuoHaoran1
(1.JiangsuKeyLaboratoryonOpto-electronicTechnology,SchoolofPhysicsScienceandTechnology,NanjingNormalUniversity,Nanjing210023,China)(2.JiangsuCenterforCollaborativeInnovationinGeographicalInformationResourceDevelopmentandApplication,Nanjing210023,China)
关键词:
无线层析成象; 无设备定位; 接收信号强度; 核主成分分析
Keywords:
radiotomographyimaging; device-freelocalization; receivedsignalstrength; kernelprincipalcomponentanalysis
分类号:
O451
DOI:
10.3969/j.issn.1001-4616.2020.01.006
文献标志码:
A
摘要:
无线层析成像(radiotomographicimaging,RTI)技术作为无设备目标定位(device-freelocalization,DFL)的主要方式之一,在被定位目标不携带任何定位装置的情况下仍能实现定位,具有广泛的应用前景.但由于接收信号强度(receivedsignalstrength,RSS)信息容易受到环境变化和噪声的影响,RTI成像图上往往不可避免地存在着背景噪点,有时甚至还有伪目标出现在图像上.为了提高RTI成像质量,本文提出一种基于核主成分分析(kernelprincipalcomponentanalysis,KPCA)的增强型RTI方法,该方法利用KPCA的学习能力来提取有效受目标影响的链路特征信息,从而达到克服噪声影响和提高定位精度的目的.室内外实验结果表明,该方法的成像质量和定位精度都要优于现有RTI方法.
Abstract:
RadioTomographyImaging(RTI)technologyisoneofthemainmethodsofdevice-freelocalization(DFL).Itcanstillachievepositioningwithoutbeingpositionedwithanypositioningdevice,andwithawiderangeofapplicationprospects.However,sincetheReceivedSignalStrength(RSS)informationiseasilyaffectedbyenvironmentalchangesandnoise,backgroundnoiseisinevitablypresentontheRTIimage,andsometimesevenpseudotargetsappearontheimage.InordertoimprovethequalityofRTI,thispaperproposesanenhancedRTImethodbasedonKernelPrincipalComponentAnalysis(KPCA),whichusesthelearningabilityofKPCAtoextractthelinkcharacteristicinformationthatiseffectivelyaffectedbythetarget,inordertoovercometheeffectsofnoiseandimprovepositioningaccuracy.TheresultsofindoorandoutdoorexperimentsshowthattheimagingqualityandpositioningaccuracyofthismethodarebetterthantheexistingRTImethod.

参考文献/References:

[1]刘凯,夏然,柴柯.结合菲涅尔理论的免携带设备定位研究[J].传感技术学报,2015,38(2):298-302.
[2]孙立奎,王洪玉,吴力飞,等.基于信号差分特性的被动入侵检测方法[J].传感技术学报,2012,25(11):1549-1555.
[3]SANGHP,SOHW,JONGBL.Smarthome-digitallyengineereddomesticlife[J].Personalandubiquitouscomputing,2003,7(3-4):189-196.
[4]李禹.UWB-TWDR的运动目标检测及定位[D].长沙:国防科学技术大学,2003,TN95:1-59.
[5]TSIMHONIO,BARGMANJ,MICHAELJ,etal.Pedestriandetectionwithnearandfarinfrarednightvisionenhancement[J].ThejournaloftheilluminatingengmeeringsocietyofNorthAmerica(Leukos),2004,4(2):113-128.
[6]BORENSTEINJ,KORENY.Noiserejectionforultrasonicsensorsinmobilerobotapplications[C]//Proceedings1992IEEEInternationalConferenceonRoboticsandAutomation.Nice,France,1992,2:1727-1732.
[7]LEDORTZN,GAINF,ZETTERBERGP,etal.WiFifingerprintindoorpositioningsystemusingprobabilitydistributioncomparison[C]//2012IEEEInternationalConferenceonAcoustics,SpeechandSignalProcessing(ICASSP).Kyoto,Japan,2012:2301-2304
[8]KOOJ,CHAH.LocalizingWiFiaccesspointsusingsignalstrength[J].IEEEcommunicationsletters,2011,15(2):187-189.
[9]PATWARIN,WILSONJ.RFsensornetworksfordevice-freelocalization:mesurements,modelsandalgorithms[J].ProceedingsofIEEE,2010,98(11):1961-1973.
[10]WILSONJ,PATWARIN.Afade-levelskew-laplacesignalstrengthmodelfordevice-freelocalizationwithwirelessnetworks[J].IEEEtransactionsonmobilecomputing,2012,11(6):947-958.
[11]LUOY,HUANGK,GUOX,etal.AhierarchicalRSSmodelforRF-baseddevice-freelocalization[J].Pervasiveandmobilecomputing,2016,31:124-136.
[12]WANGZ,LIUH,MAH,etal.Enhancingindoorradiotomographicimagingbasedoninterferencelinkelimination[J].Digitalsignalprocessing,2015,44:26-36.
[13]WILSONJ,PATWARN.Radiotomographicimagingwithwirelessnetworks[J].IEEEtransactionsonmobilecomputing,2010,9:621-632.
[14]KALTIOKALLIOO,BOCCAM,PATWARIN.Afadelevel-basedspatialmodelforradiotomographicimaging[J].IEEEtransactionsonmobilecomputing,2014,13(6):1159-1172.
[15]DAEMH,JAEHL.Designandimplementationofsmarthomeenergymanagementsystemsbasedonzigbee[J].IEEEtransactionsonconsumerelectronics,2010,56(3):1417-1425.
[16]JENNIFERY,BISWANATHM,DIPAKG.Wirelesssensornetworksurvey[J].Computernetworks,2008,52(12):2292-2330.
[17]STEPHANL,GEORGIOST,STEFANOSZ.Eulerprincipalcomponentanalysis[J].Internationaljournalofcomputervision,2013,101(3):498-518.
[18]NANL,HANW.Weightedprincipalcomponentextractionwithgeneticalgorithms[J].Appliedsoftcomputing,2012,12(2):961-974.
[19]WANGLW,WANGX,ZHANGXR,etal.Theequivalenceoftwo-dimensionalPCAtoline-basedPCA[J].Patternrecognitionletters,2005,26(1):57-60.
[20]CANNONA.Nonlinearprincipalpredictoranalysis:applicationtothelorenzsystem[J].Journalofclimate,2006,19(4):579-589.
[21]SHENGBH,WANGJL,LIP.ThecoveringnumberforsomeMercerkernelHilbertspaces[J].Journalofcomplexity,2008,24(2):241-258.

备注/Memo

备注/Memo:
收稿日期:2019-02-01.
基金项目:2018年江苏省研究生科研创新项目(KYCX18_1187).
通讯作者:柯炜,副教授,研究方向:无线电定位.E-mail:kewei@njnu.edu.cn
更新日期/Last Update: 2020-03-15