Radar Detection of Fluctuating Targets under Heavy-Tailed Clutter Using Track-Before-Detect

Area of Science:

• Radar Systems Engineering
• Signal Processing
• Statistical Detection Theory

Background:

• Heavy-tailed clutter, often modeled by K-distribution, poses challenges for conventional radar detection due to non-Gaussian noise and target fluctuations (Swerling type 1).
• Standard track-before-detect (TBD) methods exhibit performance degradation in such environments because of frequent target-like outliers.

Purpose of the Study:

• To develop and evaluate an enhanced dynamic programming track-before-detect (DP-TBD) algorithm for improved detection of fluctuating targets in heavy-tailed clutter.
• To address the limitations of conventional TBD techniques in non-Gaussian environments by incorporating prior information.

Main Methods:

• Derivation of likelihood ratio merit functions for Swerling type 1 targets under non-Gaussian backgrounds.
• Development of an efficient approximation method and a two-stage detection approach to reduce computational complexity.
• Integration of environmental and target fluctuation information into the DP-TBD process.

Main Results:

• The proposed DP-TBD algorithm demonstrates enhanced detection performance compared to conventional methods.
• Numerical simulations confirm the effectiveness of the strategy, particularly for fluctuating targets in heavy-tailed clutter.
• The approximation and two-stage detection methods successfully reduce computational load.

Conclusions:

• The novel DP-TBD algorithm offers a significant improvement in radar detection capabilities for fluctuating targets within heavy-tailed clutter.
• The proposed methods provide a computationally efficient solution for enhancing radar performance in challenging non-Gaussian environments.