Three-layerAdaptiveBack-PropagationNeuralNetworks(TABPNN)areemployedforthedemodulationofspreadspectrumsignalsinamultiple-accessenvironment.Aconfigurationemployingthree-layeradaptiveBack-propagationneuralnetworksisputforwardforthedemodulationofspread-spectrumsignalsinasynchronousGaussianchannels.Thetheoreticalargumentsandpracticalperformancebasedontheneuralnetworksareanalyzed.Theresultsshowthatwhethertheresistancetothemultipleaccessinterferenceortherobusttonear-fareffects,theproposeddetectorsignificantlyoutperformsnotonlytheconventionaldetectorbutalsotheBPneuralnetworksdetectorandiscomparabletotheoptimumdetector.
