Research On Tropospheric Scatter Signal Detection And DOA Estimation Based On Time,Frequency Domain Analysis

Maritime rights are very important rights for our country that must be controlled and maintained well.In order to effectively deal with the threat from the sea,the passive detection over the horizon(OTH)should be carried out for reconnaissance,then the target azimuth should be accurately determined.The known large offshore weapon platforms are often equipped with searching radar,whose transmit signal re-radiated by troposphere can realize to propagate OTH.So the signal detection and azimuth estimation become possible for passive detection.Firstly,the propagation loss characteristics are analyzed,especially the bias loss in the non-cooperative passive detection circumstance.The roughly received signal to Noise Ratio(SNR)is obtained combined with the typical target signal.Then,the paper analyzes the energy diffusion characteristics in time,frequency and their joint domain caused by multipath fading.The signal energy diffusion in the corresponding field is obtained by the simulation experiments,which lay a foundation for the further research on the signal detection and DOA estimation method.Secondly,according to fading diffusion characteristics in time domain,the relationship between the truncation strategy and the gain is analyzed to obtain the optimal sliding window.Using the sliding window to truncate the received signal sequence to get the energy concentrated part,the SNR can be improved.Next,the paper truncates the approximate signal spectrum range to get the gain further.Then,the energy of the truncated signal part is accumulated for detection.At last,the simulation results verify the effectiveness of the detection algorithm for the low SNR environment after the truncation and accumulation.Again,based on time-frequency MUSIC algorithm,the paper constructs the spatial time-frequency distribution matrix modified by time-frequency fading diffusion,then the SNR improvement,subspace estimation error are analyzed.Due to the increase of the eigenvalues after the eigen-decomposition,the SNR is improved and the subspace estimation variances are decreased,which in theory reflects the algorithm performance is improved.Finally,through Monte Carlo simulation experimental,the results of spatial spectral estimation and angle estimation standard deviation are analyzed to prove the algorithm performance modified by time-frequency fading diffusion treatment.Finally,a set of data acquisition and visualization software is designed based on signal detection and DOA estimation.Then,combined with the troposcatter passive detection system OTH,the paper designs a near-field test to verify the availability of the algorithm and visualization software,which lays the foundation for external-field test.