Incremental Discriminant Analysis on Interval-Valued Parameters for Emitter Identification

Abstract EN

Emitter identification has been widely recognizedas one crucial issue for communication, electronicreconnaissance, and radar intelligence analysis.

However, the measurements of emitter signal parameterstypically take the form of uncertain intervalsrather than precise values.

In addition, themeasurements are generally accumulated dynamicallyand continuously.

As a result, one imminenttask has become how to carry out discriminantanalysis of interval-valued parameters incrementallyfor emitter identification.

Existing machinelearning approaches for interval-valued data analysisare unfit for this purpose as they generally assumea uniform distribution and are usually restrictedto static data analysis.

To address the aboveproblems, we bring forward an incremental discriminantanalysis method on interval-valued parameters(IDAIP) for emitter identification.

Extensiveexperiments on both synthetic and real-life datasets have validated the efficiency and effectivenessof our method.