SCIENTIA SINICA Informationis, Volume 48, Issue 12: 1634-1650(2018) https://doi.org/10.1360/N112018-00033

A sparse estimation algorithm for the radar target range-direction under strong mainlobe jamming conditions

More info
  • ReceivedFeb 12, 2018
  • AcceptedApr 28, 2018
  • PublishedNov 27, 2018


In this paper, a sparse estimation algorithm for the radar target range and direction under strong mainlobe blanket jamming conditions is presented. This algorithm can effectively suppress multiple strong mainlobe blanket jammings (multi-MLJs) and extract the information about target distance and direction. First, the space domain adaptive processing for the received signal in each row (or column) is adopted to suppress multi-MLJs. Second, the sparse Bayesian learning is employed to obtain the information of the target distance and elevation (or azimuth). In addition, the proposed method still has an effective performance under the error condition and the applicable conditions of the new algorithm based on the performance analysis are provided in this paper.

Funded by



[1] Wang Y L, Ding Q J, Li R F. Adaptive Array Processing. Beijing: Tsinghua University Press, 2009. Google Scholar

[2] Applebaum S P, Wasiewicz R. Main Beam Jammer Cancellation for Monopulse Sensors. Final Technical Report DTIC RADC-TR-86-267, 1984. Google Scholar

[3] Yu K B, Murrow D J. Combining Sidelobe Canceller and Mainlobe Canceller for Adaptive Monopulse Radar Processing. US Patent 6 867 726, 2005. Google Scholar

[4] Rao C, Li R F, Dai L Y. Monopulse estimation with multipoint constrained adaptation in mainlobe jamming. In: Proceedings of the 6th International Conference on Radar, 2011. Google Scholar

[5] Li R F, Rao C, Dai L Y, et al. Algorithm for constrained adaptive sun-difference monopulse among sub-arrays. Huazhong Univ Sci Tech (Nat Sci Ed), 2013, 41: 6--10. Google Scholar

[6] Li R, Rao C, Dai L. Combining sum-difference and auxiliary beams for adaptive monopulse in jamming. J Syst Eng Electron, 2013, 24: 372-381 CrossRef Google Scholar

[7] Zhou B L, Li R F, Dai L Y, et al. Adaptive monopulse algorithm based on combining four-channel sum-difference beam and auxiliary elements. Syst Eng Electron, 2017, 39: 1905--1914. Google Scholar

[8] Shi Q, Wu R, Zhong L. Adaptive interference suppression based on single-channel optimal constant modulus algorithm. J Electron Inf Tech, 2011, 33: 1126-1130 CrossRef Google Scholar

[9] Dai H Y, Wang X S, Liu Y. Novel research on main-lobe jamming polarization suppression technology. Sci China Inf Sci, 2012, 55: 368-376 CrossRef Google Scholar

[10] Dai H Y, Li Y Z, Liu Y, et al. Novel research on main-lobe jamming polarization suppression technology. Sci China Inform, 2012, 42: 460--468. Google Scholar

[11] Dai H, Wang X, Li Y. Main-Lobe Jamming Suppression Method of using Spatial Polarization Characteristics of Antennas. IEEE Trans Aerosp Electron Syst, 2012, 48: 2167-2179 CrossRef ADS Google Scholar

[12] Zhang Q Y, Cao B, Wang J. Polarization filtering technique based on oblique projections. Sci China Inf Sci, 2010, 53: 1056-1066 CrossRef Google Scholar

[13] Cai Q W, Wei P, Xiao X C. Single-channel blind separation of overlapped multicomponents based on energy operator. Sci China Inf Sci, 2010, 53: 147-157 CrossRef Google Scholar

[14] Wang X, Liu J, Meng H. Novel atomic decomposition algorithm for parameter estimation of multiple superimposed Gaussian chirplets. IET Radar Sonar Navig, 2011, 5: 854-861 CrossRef Google Scholar

[15] Cai T T, Wang L. Orthogonal Matching Pursuit for Sparse Signal Recovery With Noise. IEEE Trans Inform Theor, 2011, 57: 4680-4688 CrossRef Google Scholar

[16] Lee K, Bresler Y, Junge M. Subspace methods for joint sparse recovery. Mathematics, 2012, 58: 3613--3641. Google Scholar

[17] Boyd S, Vandenberghe L. Convex Potimization. Beijing: Tsinghua University Press, 2013. Google Scholar

[18] Wipf D P, Rao B D. Sparse Bayesian Learning for Basis Selection. IEEE Trans Signal Process, 2004, 52: 2153-2164 CrossRef ADS Google Scholar

[19] Wipf D P, Rao B D. An Empirical Bayesian Strategy for Solving the Simultaneous Sparse Approximation Problem. IEEE Trans Signal Process, 2007, 55: 3704-3716 CrossRef ADS Google Scholar

[20] Tipping M E. Sparse Bayesian learning and the relevance vector machine. J Mach Learn Res, 2001, 1: 211--244. Google Scholar

[21] Ding L, Li R, Wang Y. Discrimination and identification between mainlobe repeater jamming and target echo by basis pursuit. IET radar sonar navigation, 2017, 11: 11-20 CrossRef Google Scholar

[22] Ding L, Li R, Dai L. Discrimination and identification between mainlobe repeater jamming and target echo via sparse recovery. IET radar sonar navigation, 2017, 11: 235-242 CrossRef Google Scholar

  • Table 1   The simulation parameters setting
    Parameter Setting
    The array antenna Consider a rectangular planar array which has 24 columns, and each column
    has 20 elements.
    The element spacing Half a wavelength.
    The beam boresight $\left(~\text{9}0{}^\circ~,30{}^\circ~~\right)$, the former is azimuth, and the latter is elevation.
    The 3 dB beam width $\left(~4.21{}^\circ~,5.05{}^\circ~~\right)$.
    The transmission signal Linear frequency modulation (LFM) signal, with bandwidth $B=5$ MHz,
    pulse width $\tau~=20~\mu$s and sampling rate ${{f}_{s}}=10$ MHz.
    The target One target with 5 dB is located at $\left(~\text{9}0{}^\circ~,30{}^\circ~~\right)$ and the 500-th sampling point.
    The jammings Two mainlobe blanket jammings with 50 dB are located at $\left(~\text{91}.05{}^\circ~,31.26{}^\circ~~\right)$,
    $\left(~88.95{}^\circ~,28.74{}^\circ~~\right)$, which are located at the one quarter of 3 dB beam width.
    Noise White Gaussian noise with 0 dB.
    The observation sample 5.
    number of sparse recovery

    Algorithm 1 Algorithm

    Inputs: the array received signal ${\boldsymbol~x}\in~{{\mathbb{C}}^{{{N}_{1}}\times~{{N}_{2}}}}$, angle dictionary $\mathbf{\varphi~}\in~{{\mathbb{C}}^{{{N}_{1}}\times~L}}$. Outputs: the sparsity coefficient $\mathbf{\omega~}$. 1. The outputs by the spatial row adaptive processing: ${\boldsymbol~r}(~n~)={\boldsymbol~a}\left(~{{\varphi~}_{0}}~\right)p{{s}_{0}}(~n~)+{{{\boldsymbol~v}}_{r}}(~n~)$; 2. Selecting $K$ training samples: ${\boldsymbol~r}\in~{{\mathbb{C}}^{{{N}_{1}}\times~K}}$; 3. Setting initial values: ${{\mathbf{\gamma~}}_{0}}=1$, $\sigma~_{0}^{2}={{10}^{-1}}$; 4. Calculating the posterior component:

Copyright 2020 Science China Press Co., Ltd. 《中国科学》杂志社有限责任公司 版权所有