,nippression ofNarrowaIand' Interference 'in DSSSSystems EasedonMAWPTandCWPT
Mingyan Jiang, Dongfeng Yuan, Guoping Xu · 2005
Two new NarrowBandInterference (NBI) (WPT)(1,2).Because ofthelimitation ofFFT,theNBIcannot suppression methods basedonMiuti-band Wavelet Packet beexcise thoroughly inthealgorithms based onFFT.Because Transform (MWPT)andComplex Wavelet Packet TransformWPThasgoodcharacter ofmiutiresolution analysis ability and (CWPT)inDirect Sequence Spread Spectrum (DSSS) systemstime-frequency localization, itcanexactly trackandlocate arepresented inthis paper. Withthese twotechniques, jammed theposition oftheNBI inthefrequency domain, andthen DSSSsignal isdecomposed separately using MWPTandCWPT effectively filter NBIoutbyusingtreestructure algorithm. inorder toseparate desirable signal fromjamming signal. The Inourpaper, we adoptMWPT andCWPT intheNBI DSSSsignal isrecovered bywavelet reconstruction after the suppression technology inDSSSsystem, thebiterror ratio interference signal isfiltered out. Theresults ofoursimulations (BER)ofouralgorithms outperform theoldalgorithm. showthat these twotechniques outperform thescheme based on 2-band wavelet packet transform. 2.MWPT andCWPT 2.1MWPT M-band wavelet isoneimportant branch ofwavelet theory 1.SS coIuntrou cation system has the inherent capabihty of developed inrecent years. Itssymmetric, smooth andcompactly DSSScommunication system hastheinherent capability of supported properties arebetter thian 2-band wavelet's, soM-band rejecting thedeliberate orunintentional interference. InDSSS wavelet hasthepossibility tosolve theproblems that 2-band communication, the increase inbandwidth yields aprocessingwavelet can't do. gain that provides desired interference suppression capability. wavelet can't do. However, thefrequency resource islimited, whenhigh-power SetM > 2 themultiscale analysis ofsquare-integral interference appears, only magnifying theprocess gain ofDSSS vector spaceL2(R)isasequence {V} ofclose subspace isnotenough. Theemploying ofinterference suppression in DSSSsystem becomes necessary. satisfying certain conditions. Assumethatqp(x) and Thetechnology ofinterference suppression canbedivided into twocategories: oneisestimating filtering intimedomain f (x),l .s.M - 1}arethecorresponding scale function (Wiener filtering). Thetransverse filter canonly beusedin slowchanging interference suppression(l). Theother is andwavelet function, then wecangettwofamilies offunctions processing methods intransform domain, these kinds of bydilating andshifting theoriginal finctions: methods neednoadaptive algorithm, thenthey cansuppress 5Dk(x) =MA 2q(M-Jx-k) (l) interference inreal time. Thetransforms frequently employed ininterference suppression inDSSSsystem are:FFT', f (x)=M '2gs(Mx j- k) (2) traditional 2-band waveletpacket tranlsform