Abstract:Tosolvetheproblemofthemismatcherrorsbetweentheactualandpresumedsignalsteeringvectorsinairborneradarspace-timeadaptiveprocessing(STAP)

arobustreduced-dimensionSTAPmethodbasedonbi-iterativealgorithm(BIA)andsecond-orderconeprogramming(SOCP)isproposed.FirstlytheweightvectorisdecomposedintotheKroneckerproductofthespatialandtemporalweightvectors.Thenbyimposingaboundontheerrorsbetweentheactualspatialandtemporalsteeringvectorsandthepresumedsteeringvectors

respectively

theoptimizationoftheworstcaseperformanceisexecutedandtransformedintotheSOCPformulation.FurthertheBIAisutilizedtoresolvethetwolowdimensionweightvectors

andfinallythefulldimensionSTAPweightvectorissynthesized.TheproposedmethodcansignificantlydecreasethetrainingsamplerequirementandthecomputationalcomplexitywhilemaintainingtherobustSTAPperformance

whichismorevaluableforpracticalengineeringapplication.Simulationresultsverifytheeffectivenessoftheproposedmethod.