Incoherent Integration Of Radar Signals In Range Doppler Domain And Its DSP Implementatio

Target detection is a hotspot in the radar field.Weak targets are difficult to be detected because its small Radar Cross-Section and the long radial distance.Long-term integration to the energy can effectively improve the signal-to-noise ratio of the echo signal.An innovative weak target detection method is proposed in this thesis,which is based on target inter-frame correlation and noise characteristics,linearly filtering the data matrix of moving target detection by discrete Kalman filter and Rao-Blackwellized particle filter.The DSP implementation of this method is completed in the radar system.First of all,the important models in radar target detection and key technologies in traditional radar signal processing are introduced in this thesis.We also introduce the weak target detection technology based on coherent integration of single-frame and non-coherent integration of multi-frame.Secondly,we propose a multi-frame signal accumulation method based on Kalman filter in the Range-Doppler domain.It directly filters the MTD data matrix to efficiently improve the signal-to-noise ratio compared with existing methods according to the difference between the noise and target echo.Thirdly,the previous method requires noise parameters,so we propose another method based on Rao-Blackwellized particle filter.We first estimate the required parameters based on particle filter,and then improve the signal-to-noise ratio based on Kalman filter.Two methods are proved by MATLAB simulation under different conditions.Finally,the project implementation based on DSP has been completed based on the project of semi-active/active composite radar.The test result proves that the method has certain practical application value.