Ultra Wideband Noise SAR Imaging Based On Compressed Sensing

Radar imaging technology has been widely used in many scientific fields,such as ground target detection,earth environment mapping,oceanography and glacier research.The imaging algorithm mainly involve matched filtering process,this requires the signal receiver to operate at the sampling frequency determined by the Nyguister sampling law.In order to further improve the imaging resolution,it is necessary to use higher sampling rate signal acquisition equipment.At this stage,it is very difficult to increase the sampling rate of radar signal acquisition equipment by a large margin,meanwhile the large sample data generated by the higher signal acquisition devices will result in only fewer target scenarios can be measured for the limited storage devices.Compressed sensing(CS)theory appeared around 2006 and has been successfully applied to many fields by many scholars.CS theory shows that using prior knowledge of the signal,only a small set of measurement data is needed to identify and reconstruct it.When applied to high resolution radar imaging,the number of signal samples can be reduced effectively,and the sampling rate of signal acquisition devices can be significantly decreased.In this paper,the radar imaging using ultra-wideband(UWB)noise as signal source is introduced.Compared with traditional radar systems and imaging algorithm,it mainly contains the following advantages.First,the random characteristics of the signal source make the measurement less interceptable and the system has strong anti-electromagnetic interference property.Second,when using the same amount of sampling data,the image quality can be improved by using CS based method,and the side-lobe masking effect of the traditional imaging algorithms can be eliminated.Third,utilizing the sparsity of imaging scene,it can be reconstructed accurately with signal down sampling,even if the sampling frequency is lower than the Nygurster sampling law.Therefore,high resolution image can be obtained by using lower sampling equipment.Fourth,since the sensing matrix of CS theory needs to satisfy certain conditions,when using the determined signal as the signal source,random sampling method is required to ensure this property.However,the random characteristics of UWB noise sources can make the sensing matrix meet this requirements even without random sampling,so the system can be simplified.The main contents of this paper are as follows:(1)This paper summarizes the history of radar imaging theroise,as well as the background of various imaging algorithms,and focuses on the present situation of CS based method like the opportunities or challenges of the application for UWB noise radar.(2)This paper briefly introduces the traditional imaging theory of UWB noise radar,as well as the basic principle of CS theory and its solving algorithm.(3)An approach of UWB noise radar imaging based on CS is introduced in this paper,which can eliminates the side-lobe masking effect of strong reflection points compared with traditional imaging algorithms.Sampling data utilization can be reduced to 36% without affecting the final image quality,thus reducing the system cost.(4)For the complexity of the CS reconstruction algorithm,a new fast reconstruction algorithm is introduced,which has good computational efficiency and accuracy,and can be used for large-scale scenes imaging.The imaging with 2048×2048 pixels was finally achieved,resulting in a 256 fold improvement for the image size.