Fast Implementation of Approximated Maximum Likelihood Parameter Estimation for Frequency Agile Radar under Jamming Environment

Yang Zhao, Jianxin Wu, Zhiyong Suo, Xiaoyu LiuSensors 2020 · DOI

TL;DR

This paper presents a computationally efficient algorithm for estimating target parameters in Frequency Agile Radar (FAR) systems operating in challenging jamming environments. By combining adaptive beamforming for barrage noise suppression and frequency agility for deceptive jamming mitigation, followed by a novel low-order approximation to the Maximum Likelihood (ML) estimation, the authors significantly reduce computational complexity compared to traditional fine grid-search methods, achieving a good balance between performance and efficiency.

What Problem Does This Paper Address?

Modern radar systems, particularly Frequency Agile Radars (FARs), face increasingly complex electromagnetic environments due to sophisticated jamming techniques. These include barrage noise jamming, which floods the radar with broadband noise, and deceptive jamming, which aims to mislead the radar about target characteristics. Accurately estimating target parameters (like range, velocity, and angle) is crucial for effective target identification and tracking. However, traditional Maximum Likelihood (ML) estimation, while optimal in theory, suffers from extremely high computational complexity, especially in multi-dimensional parameter spaces and under jamming conditions, making it impractical for real-time applications. Existing methods often struggle to balance jamming suppression, parameter estimation accuracy, and computational feasibility.

Methods

The proposed method tackles this problem in two main stages:

  1. Jamming Suppression:
    • Barrage Noise Jamming: Addressed using adaptive beamforming techniques.
  2. Deceptive Jamming: Mitigated by leveraging the inherent frequency agility of the FAR. The paper notes that deceptive jamming is often suppressed by matched filtering, and the frequency agility further enhances this capability.
  3. Efficient Parameter Estimation:
    • Approximated Maximum Likelihood (ML): Instead of performing a computationally intensive fine grid-search over the multi-dimensional parameter space (range, velocity, angle), the authors derive an analytical solution. This is achieved by applying a low-order Taylor approximation to the multi-dimensional ML function.
  4. Parameter-Independent Weight Vectors: To further simplify the ML solution, the authors introduce parameter-independent adaptive sum and difference weight vectors. This approximation avoids the need to search for optimal weight vectors at each step, significantly reducing computational load.
  5. Iterative Refinement: An iterative technique is employed to improve the accuracy of the Taylor approximation, using estimated target parameters from the previous iteration.

Datasets

The paper does not introduce new datasets. Instead, it relies on simulated data to validate the proposed algorithm. The simulations are designed to mimic realistic scenarios with multiple targets and various types of jamming.

Key Insights and Ideas

  • Frequency Agility as a Jamming Countermeasure: The paper highlights that frequency agility is not just for improving range-velocity resolution but is a powerful tool for actively suppressing deceptive jamming.
  • Approximation for Computational Efficiency: The core innovation lies in approximating the complex ML function. By using a low-order Taylor expansion and parameter-independent weight vectors, the authors transform a computationally intractable problem into a solvable system of equations.
  • Trade-off Between Complexity and Accuracy: The proposed method demonstrates that a significant reduction in computational complexity can be achieved with only a "trivial performance loss," suggesting a favorable trade-off for practical radar systems.
  • Balancing Act: The authors position their method as achieving a good balance between computational cost and target parameter estimation performance, outperforming simpler adaptive methods while being far more efficient than brute-force grid search.

Key Results

  • Computational Complexity Reduction: The paper shows a significant reduction in computational complexity compared to the fine grid-search (FGS) method. For instance, with 40 array elements, the proposed ML method has a complexity of approximately $10^{11}$ operations, while FGS is around $10^{14}$ operations.
  • Jamming Suppression Effectiveness: Simulations demonstrate that the frequency agile case effectively suppresses deceptive jamming, preserving target signals while attenuating jamming signals, as shown in Figures 2, 3, and 4.
  • Performance in Various Conditions:
    • SNR: The proposed ML method's performance (measured by Normalized Root Mean Square Error - NRMSE) approaches that of the FGS method across various Signal-to-Noise Ratio (SNR) levels, outperforming the adaptive monopulse method.
  • Jamming: As the number of barrage noise jamming increases, the RMSE for all methods increases, but the proposed ML method maintains a competitive performance, closely following FGS and outperforming adaptive monopulse.
  • Angle Deviation: The method shows robust performance even when the target angle deviates from the line of sight, achieving a good balance between FGS and adaptive monopulse.
  • Line of Sight: Similar to angle deviation, the method demonstrates reasonable performance as the line of sight varies, again finding a middle ground between FGS and adaptive monopulse.

Limitations and Future Work

  • Approximation Accuracy: While the paper claims "trivial performance loss," the approximation inherently introduces some error. The extent of this loss might be more significant in extreme jamming scenarios or with very complex target dynamics not fully captured by the simplified model.
  • Model Assumptions: The method relies on specific models for jamming and target echoes. Its performance might degrade if these assumptions are significantly violated in real-world scenarios.
  • Future Work: The authors suggest that the presented method achieves a good balance between computational cost and estimation performance. Further research could explore extending this approximation to even more complex scenarios or investigating adaptive ways to adjust the approximation order based on the environment.

Summary updated

Abstract

A computationally efficient target parameter estimation algorithm for frequency agile radar (FAR) under jamming environment is developed. First, the barrage noise jamming and the deceptive jamming are suppressed by using adaptive beamforming and frequency agility. Second, the analytical solution of the parameter estimation is obtained by a low-order approximation to the multi-dimensional maximum likelihood (ML) function. Due to that, fine grid-search (FGS) is avoided and the computational complexity is greatly reduced.
BibTeX
@inproceedings{zhao2020fast,
  author = {Yang Zhao and Jianxin Wu and Zhiyong Suo and Xiaoyu Liu},
  title = {{Fast Implementation of Approximated Maximum Likelihood Parameter Estimation for Frequency Agile Radar under Jamming Environment}},
  booktitle = {Sensors},
  year = {2020},
  doi = {10.3390/s20072022},
}
Figure