KWT-Based Fast Target Motion Parameter Estimation in Three-Channel SAR-GMTI

Literature Overview

The paper by Qian Jiang, Su Junhai, Li Lianghai, and Xing Mengdao, published in Journal of Electronics & Information Technology in 2010, presents a method for estimating the motion parameters of fast-moving targets in airborne three-channel SAR-GMTI systems using the Keystone-Wigner Transform (KWT). This work builds upon the authors' earlier contributions to SAR-GMTI processing and addresses the specific challenge of accurate velocity estimation when targets exhibit both radial and along-track motion components. The research is supported by the National Natural Science Foundation of China (Grants 60890072 and 60802081) and represents a significant advancement in GMTI signal processing methodology.

Problem Formulation

Dual Doppler Shift Components

In airborne SAR-GMTI systems, a moving ground target's Doppler frequency shift contains two components:

• Radial velocity component: Causes a constant Doppler shift proportional to the target's radial velocity relative to the radar. This shifts the target's Doppler center away from the clutter notch.
• Along-track velocity component: Causes a change in the azimuth chirp rate, leading to azimuth defocus in conventional SAR imagery. The along-track velocity modifies the effective platform velocity relative to the target, altering the range-Doppler relationship.

Ambiguity Challenges

For fast-moving targets, two ambiguity problems arise simultaneously:

1. PRF ambiguity: When the radial velocity is high, the Doppler frequency may exceed the PRF, causing spectral folding. A measured Doppler frequency corresponds to multiple possible true velocities.
2. Interference phase ambiguity: The interferometric phase measured between channels is periodic, meaning that the measured phase corresponds to multiple possible radial velocities differing by integer multiples of the phase ambiguity interval.

Proposed Methodology

Overall Processing Flow

The proposed method follows a structured processing chain:

1. Dechirp-domain target detection: Moving target detection is performed in the dechirp domain, where the target signal appears as a frequency-modulated signal with parameters related to target motion.
2. Radial velocity estimation with deambiguity: The target's radial velocity is estimated by combining the interferometric phase measurement with the range migration rate information to resolve both PRF ambiguity and phase ambiguity.
3. Along-track velocity estimation via KWT: After compensating for range migration, the target signal is extracted and analyzed using the Keystone-Wigner Transform to determine the along-track velocity component.
4. Final velocity synthesis: The radial and along-track velocity components are combined to obtain the complete target velocity vector.

Keystone-Wigner Transform

The Keystone Transform (KT) is a well-established technique for range migration compensation in SAR processing. It warps the range-frequency data such that all range migration effects are removed, enabling azimuth focusing through conventional Fourier transform processing. The Wigner Distribution (WD) is a time-frequency analysis tool that provides joint time-frequency representation of a signal.

The Keystone-Wigner Transform (KWT) combines these two techniques: after applying the Keystone Transform to compensate for range migration, the Wigner Distribution is computed to extract the chirp rate of the target signal. The chirp rate directly corresponds to the along-track velocity component of the target.

PRF and Phase Deambiguity

The deambiguity strategy is a key innovation of this work. The range migration rate, which can be measured from the target's trajectory in the range-Doppler plane, provides information about the radial velocity that is independent of the Doppler frequency measurement. By combining the migration rate information with the interferometric phase measurement, the algorithm can resolve the PRF ambiguity and determine the true radial velocity.

The interferometric phase between two channels separated by a baseline d is related to the radial velocity v_r by:

φ = 4π · d · v_r / (λ · v_p)

where λ is the radar wavelength and v_p is the platform velocity. Since the phase is periodic with 2π, the measured phase φ_m corresponds to:

v_r = (φ_m + 2πn) · λ · v_p / (4π · d)

where n is an unknown integer. The range migration rate provides an independent estimate of v_r that can be used to determine the correct value of n.

Technical Parameters and Performance

System Parameters (Typical)

Performance Characteristics

• Radial velocity accuracy: Sub-meter per second accuracy achievable with proper phase measurement and deambiguity.
• Along-track velocity accuracy: Dependent on the quality of the Wigner Distribution estimation; typically accurate to within 1-2 m/s for targets with sufficient signal-to-noise ratio.
• Maximum unambiguous radial velocity: Determined by the PRF and the deambiguity strategy; can be extended well beyond the nominal PRF limit.
• Minimum detectable velocity: Limited by the clutter suppression performance and the sensitivity of the interferometric phase measurement.

Engineering Relevance

Comparison with Alternative Methods

Practical Implementation Considerations

1. Computational complexity: The Wigner Distribution computation is computationally intensive (O(N²) for a signal of length N). Fast Wigner Distribution algorithms or approximations may be necessary for real-time implementation.
2. Phase noise sensitivity: The interferometric phase measurement is sensitive to phase noise from the radar system. Phase noise compensation techniques should be applied before phase measurement.
3. Clutter suppression quality: The performance of the entire processing chain depends on effective clutter suppression. Residual clutter can mask weak targets and introduce false detections.
4. Multiple target handling: The method assumes a single dominant target in the processing cell. Multi-target scenarios require additional processing to separate individual target signatures.

Study Insights

This work represents a significant methodological advancement in GMTI signal processing by integrating the Keystone Transform and Wigner Distribution into a unified framework for fast target velocity estimation. The key insight is that the range migration rate, which is typically treated as a nuisance parameter requiring compensation, can also serve as a useful observable for velocity deambiguity. This dual use of the migration rate information is elegant and effective.

The practical value of this method lies in its ability to provide accurate velocity estimates for targets moving at high speeds, which are precisely the targets that are most difficult to detect and track with conventional GMTI systems. In surveillance and reconnaissance applications, the ability to accurately measure the velocity of fast-moving vehicles is critical for target classification and threat assessment.

The field data results presented in the paper validate the theoretical framework and demonstrate that the proposed method achieves accurate velocity estimation in real-world conditions. This validation is essential for building confidence in the method's practical applicability.