Research On Weak Target Detection Based On Distributed Network Radar System In Complex Environment

In recent years,frequent applications of small Unmanned Aerial Vehicles(UAVs)has raised the importance of protection of securities and privacies against low,small and slow(LSS)objects,via detection and tracking of those targets.This detection and tracking process can be realized by the implementation of a distributed network radar system.With a distributed setup of multiple small radar nodes,the echoes are fused at the signal level as well as the position level,leading to tracking and identification of weak targets.In this thesis,research is conducted with focus on signal-level fusion of distributed network radar systems.Firstly,the technical background and developments of distributed network radars are presented,their basic principles and working modes analyzed.The signal model and the conditions for full-coherence operation are derived.Then,aiming on the key problems for signal-level fusion,a coherent parameter estimation methodology is proposed based on orthogonal frequency division linear frequency modulation(OFD-LFM)signal and phase coded signal.The effects of coherent parameter estimation errors on the synthesis performance are analyzed.Finally,the problem of coherent parameters estimation error,which is the core concern of the coherent synthesis system of distributed network radar,is addressed with special emphasis.The specific work of this paper is as follows:1.The coherent synthesis system of a distributed network radar composed of multiple small node radars is introduced.Its signal processing framework and working flow is given in detail.A mathematical model of radar echo signal is derived and a suitable orthogonal waveform for the application is designed.The principle of spatial energy interference of radar signals at the area of the target is analyzed and the conditions for transmission coherency of the system are given.2.Firstly,simulations and analyses are performed regarding the performance of the OFD linear modulated signal transmitted by radar nodes,to ensure that the distributed network radar works under a proper condition.Then,the simulation of the receive-coherent mode and fully coherent mode of the two-node distributed network radar system is performed.The simulation demonstrates that 6 dB signal to noise ratio(SNR)gain can be obtained by receive-coherent fusion and 9 dB SNR gain can be obtained through fully coherent fusion.Finally,the performance comparison between two coherent parameters estimation methods,peak-picking method and cross-correlation method,is given and the impact of the accuracy of coherent parameters estimation on the coherent synthesis performance is also analyzed.3.The influence of signal model parameters and input SNR of receivers on the accuracy of coherent parameters estimation is analyzed under the condition of ideal separation and non-ideal separation of transmitting signals.The influence of target scattering characteristics,including target radar cross section and multi-scattering center model,on the accuracy of coherent parameters estimation is analyzed.