Sequence set design for waveform-agile coherent radar systems

Jindong Zhang, Naiqing Xu, Song Hu, Chao ZhangEURASIP 2020 · DOI

Abstract

With increased degrees of freedom of the transmitter, coherent waveform-agile radar system can change its transmission on-the-fly in response to target detection’s requirement. This approach can provide better performance than a single waveform. In this paper, we consider unimodular sequence set design (USSD) problem based on summed range-Doppler ambiguity function (SRDAF) for coherent waveform-agile radar system. We present two algorithms for constructing unimodular sequence sets under the constant module constraint with desired minimized sidelobes on a predefined area of range-Doppler plane. The proposed algorithms are constructed based on singular value decomposition (SVD) and cyclic algorithm (CA), respectively. Numerical examples show the effectiveness of the proposed algorithms.

Introduction

Modern radar systems are increasingly moving towards "waveform agility," where the transmitter can adapt its signal pulse-by-pulse. This flexibility allows for significantly improved performance compared to traditional systems that rely on a single, static waveform. However, this shift introduces a complex design challenge: how to optimize an entire set of sequences to work together coherently. The primary goal is to minimize interfering "sidelobes" in the range-Doppler plane, which can otherwise mask small targets in the presence of larger ones or heavy clutter.

Problem Formulation and Methodology

The research addresses the Unimodular Sequence Set Design (USSD) problem. A critical constraint in radar hardware is the need for "unimodular" sequences—signals with constant amplitude where only the phase varies—to ensure maximum power efficiency.

The authors focus on optimizing the Summed Range-Doppler Ambiguity Function (SRDAF). By considering the aggregate response of the entire pulse train rather than individual pulses, the system gains additional degrees of freedom to suppress sidelobes in specific regions of interest. The resulting optimization problem is a difficult "quartic" (fourth-degree) form, which the authors simplify into a more manageable quadratic form to develop two primary algorithms:

  1. SVD-USSD (Singular Value Decomposition): This method utilizes SVD to identify optimal sequences. While it converges in fewer iterations, it is computationally intensive ($O(M^3)$ complexity) due to the required matrix decompositions.
  2. CA-USSD (Cyclic Algorithm): Designed as a lower-complexity alternative, this algorithm avoids heavy matrix mathematics, offering significantly faster performance per iteration, making it suitable for real-time applications or longer sequence sets.

Both algorithms employ an "alternating direction" approach, iteratively optimizing variables to suppress sidelobes within a predefined range-Doppler area.

Experimental Results

The proposed algorithms were evaluated through extensive numerical simulations, using a pulse train of $N=64$ pulses and sequence lengths of $M=100$ subpulses. Key findings include:

  • Sidelobe Suppression: Both SVD-USSD and CA-USSD achieved deep sidelobe suppression of approximately -50 dB in targeted regions. This represents a significant improvement over standard benchmarks like Golomb sequences or random sequences.
  • Comparison with Coordinate Descent (CD): The proposed methods outperformed existing CD algorithms (e.g., CD-64PSK). While CD methods often restrict phases to discrete sets, the SVD and CA approaches allow for continuous phase values, providing the flexibility needed for a lower sidelobe floor.
  • Target Resolution: In scenarios with multiple closely spaced targets, the optimized sequences successfully resolved all five scattering points. In contrast, standard sequences produced sidelobes that obscured smaller targets.
  • Computational Efficiency: For larger sets (e.g., $N=128$), CA-USSD demonstrated its speed advantage, completing in ~252 seconds compared to ~1382 seconds for SVD-USSD.

Conclusion and Future Work

This paper demonstrates that optimizing sequence sets via SRDAF is a powerful approach for enhancing radar "vision." While SVD-USSD offers high precision, CA-USSD provides a computationally efficient path for larger-scale problems.

Future research will look toward extending these methods to MIMO (Multiple-Input Multiple-Output) radar systems and improving real-time adaptation speeds to the millisecond level. Additionally, adapting these algorithms for more complex, non-linear clutter environments remains an open area for investigation.

Summary updated

Abstract

With increased degrees of freedom of the transmitter, coherent waveform-agile radar system can change its transmission on-the-fly in response to target detection’s requirement. This approach can provide better performance than a single waveform. In this paper, we consider unimodular sequence set design (USSD) problem based on summed range-Doppler ambiguity function (SRDAF) for coherent waveform-agile radar system. We present two algorithms for constructing unimodular sequence sets under the constant module constraint with desired minimized sidelobes on a predefined area of range-Doppler plane. The proposed algorithms are constructed based on singular value decomposition (SVD) and cyclic algorithm (CA), respectively. Numerical examples show the effectiveness of the proposed algorithms.
BibTeX
@inproceedings{zhang2020sequence,
  author = {Jindong Zhang and Naiqing Xu and Song Hu and Chao Zhang},
  title = {{Sequence set design for waveform-agile coherent radar systems}},
  booktitle = {EURASIP},
  year = {2020},
  doi = {10.1186/s13634-020-00689-0},
}
Figure