Feature Extraction And Target Detection From Complex Radar Echo

“Intelligent” technique considered to be a core point for the next generation radarsystems has played a more and more important role in technology progress and defenseapplications. For radar signal processing, how to apply characteristic analysis andextraction followed by target detection for weak targets in the complex radar returnsignals is a desirable investigation topic with both theoretic and real values. By learningthe dynamic clutter environments characteristics and analyzing the radar return signals,a higher signal-to-clutter ratio and a lower false rate can be achieved with effectivecharacteristic extraction methods and proper targets detection approaches.This dissertation is based on the existing work on complex data analysis and radarsignal processing and focuses on the investigation of characteristic extractionapproaches for the complex radar signals. Then the measured data are used to illustratethe effectiveness of the proposed approaches for weak targets detection. In addition, thepaper also describes the framework for the system analysis of the complex radar signals.The main contents are listed as follows:1) Considers the characteristics analysis methods of the complex radar signals. Itfirst takes the quantificational analysis to the complexity and uncertainty of the signals,then introduce the measured data which are used by following contents. A pretreatmentmethod based on Scaled Correlation Analysis is proposed. Discussion on the analysis ofcomplex radar echo series is finally extended to joint-domain, and the concrete methodsof multi-component non-stationary signal are investigated under both broadband andhigh-order circumstances.2) Several feature extraction methods are investigated in complex radar echo.based on topological pattern discovery. This approach reconstructs the phase space ofcomplex radar signals by extracting the topological structural features and analyzing thetopological invariants.3) Two targets detection methods in the complex radar signals is proposed,including weak targets detection based on micro-Doppler analysis, weak targetsdetection based on topological structural features, constant false alarm rate detectionbased on3D feature space of radar signal. Numerical results validate the performance ofthe proposed methods in very low SCR environments.To sum up, based on investigation background of complex environment and target,the dissertation and presents some contribution for future works. Around the keyquestions of characteristic analysis, feature extraction and weak target detection, theresearch provide theoretic summary and real applications in radar signal processing.