UsingTime-VaryingAR(TVAR)modelandadaptivenotchfilterisanewmethodforthenon-stationaryjammersuppressioninDirectSequenceSpreadSpectrum(DSSS).TheperformanceofTVARmodelforInstantaneousFrequency(IF)estimationwillbeaffectedbysomefactorssuchasbasisfunctions.Focusingonthisproblem,theoptimalbasisfunctionofTVARmodelfortheIFestimationoftheLFMsignalisobtainedinthispaper.Besidesthedepthandwidthofnotching,thephasepropertiesofnotchfilteraffecttheSignal-to-Interferenceplus-NoiseRatio(SINR)ofcorrelationoutputtothenarrowbandjammersuppressioninDSSS,inresponsetotheproblemtheclosedsolutionofcorrelationoutputSINRimprovementhasbeenderivedwhenasinglefrequencyjammerpassesthroughdirectIIRnotchfilter,anditsperformancehasbeencomparedwiththoseoffivecoefficientFIRfilters.Later,anovelmethodforLFMjammersuppressionbasedonFourierbasisTVARmodelanddirectIIRnotchfilterisproposed.Thesimulationresultsshowtheeffectivenessoftheproposedmethod.
