This proposal describes a hybrid detection, classification, and localization (DCL) system capable of detecting, classifying, and localizing radio frequency (RF) emitters in challenging scenarios involving limited visibility, measurement error, limited pulse observations, dense RF background, ambiguous emitters, and unknown/out-of-library emitters. The remote nodes are centrally controlled to acquire maximally informative measurements subject to constraints on instantaneous bandwidth and data communication/processing rates.sensor network,detection,classification,localization,Bandwidth Allocation,Exponentially Embedded Family (EEF),Bayesian inference,feature selection