Radio Frequency Interference Suppression Methods On High-frequency Surface Wave Radar

Recently,due to the low attenuation characteristics of highly conductive ocean surface with respect to vertically polarized electromagnetic waves in High-Frequency(HF)band(3-30 MHz),high-frequency surface wave radar(HFSWR)has been widely applied in over-thehorizon measurement for ocean surface dynamics parameters(wind,wave and current filed)and moving object.The frequency occupancy rate of HF band is very high,and there are a large number of active electromagnetic signals from other external radio transmitting devices in free space,which will cause serious radio Frequency interference(RFI)after entering the radar receiver and greatly reduce the radar data quality and utilization.The paper starts works from HFSWR signal processing flow and deduces the RFI characteristics in different domains by combining with the mathematical model of RFI signal.Then the paper proposes two novel RFI suppression methods on HFSWR and current results inverted from long-time measured data are used to evaluate their performance.The main works of the full text are as follow:1)This paper analyzes the sources,multi-domain characteristics and main hazards of RFI on HFSWR.Firstly,the transmitted and received signal model and signal processing flow on linear frequency modulated Interrupted continuous wave(FMICW)HFSWR are introduced.Based on the single-frequency signal model of RFI,the mathematical expressions in each signal processing stage after RFI signal enters the receiver are derived,and the RFI characteristics in time domain,frequency domain and space domain are summaried.Then combined with the single station radial current inversion process on HFSWR,the main harm of typical RFI to the ocean current inversion results is analyzed.2)The time-domain RFI suppression method based on compressed sensing(CS)is studied in this paper.The basic principle of CS theory is introduced.Based on the short-time characteristic in raw time domain and the sparsity of ocean echo in range domain of RFI,a new time-domain RFI suppression method based CS is proposed.Combined with the practical situation that range bins of the sea echo on HFSWR are known,the sparse reconstruction algorithm called orthogonal matching persuit is simplified and computation complexity is reduced because of omitting the process of iterative atom selection.The method is used to process the simulated RFI and measured RFI on measured data from OSMAR array HFSWR.The results show that the proposed CS method achieve good performance on suppressing the RFI which shows the short-time characteristic.3)A higher-order singular value decomposition(HOSVD)based multi-domain subspace projection method to suppress RFI is proposed.The conventional SVD-based RFI suppression methods,frequency-domain subspace projection(FSP)method and azimuth subspace projection(ASP)mthod,are introduced.An HOSVD-based multi-domain subspace projection method is proposed to suppress RFI.Simulations are conduceted,the results show that the HOSVD method can achieve better perforamance than the FSP method on preventing sea echoes from being attenuated;compare with the ASP method,the HOSVD method has a significant advantage that its performance is less affected by the array size and the desired signals can be kept intact.This paper compares the radial current inversion results before and after RFI mitigation of long-time measured data for 43 hours.The results show that the inverted currents after RFI suppression follow the physical characteristics of currents in Taiwan Strait and the non-physical characteristics of original current data have been eliminated.It further verifies the stability and practicability of the proposed HOSVD method.