Nearfield target localization with a few snapshots for sonar array

Zhuo Li, Deliang Liu, Xiwei Guo, Peng HeEURASIP 2018 · DOI

Abstract

Sonars can be used for detection of targets located on the sea floor where the signal undergoes multipath propagations which may give rise to large localization errors. Besides, underwater environments are often highly unstable, meaning that a large number of snapshots—requiring long sampling times—are not reliable. In sonar systems, wideband chirp signals are typically used. This paper presents a nearfield target localization method for underwater environments. First, the time of arrival (TOA) and the direction of arrival (DOA) for nearfield wideband chirp signals are estimated with only a few snapshots by using the fractional Fourier transform (FRFT) and the SPICE algorithm. Then, a virtual station method is used so that the multipath problems can be converted into line-of-sight (LOS) problems, and its Cramer-Rao lower bound (CRLB) is analyzed. Finally, simulation results demonstrate the effectiveness of the proposed method.

Introduction

Underwater target localization is notoriously difficult due to two main factors: multipath propagation and environmental instability. In shallow water or seafloor environments, sonar signals bounce off the surface and bottom. These reflections (multipath) create "ghost" signals that lead to massive localization errors if treated as simple line-of-sight (LOS) paths. Furthermore, the underwater medium is highly dynamic—currents, moving targets, and biological noise mean that the environment changes rapidly.

Traditional high-resolution algorithms like MUSIC or ESPRIT require a large number of "snapshots" (samples over time) to work effectively. However, in an unstable ocean, taking many snapshots is unreliable because the target or the medium may have shifted during the sampling window. This research addresses the need for high-resolution Direction of Arrival (DOA) and Time of Arrival (TOA) estimates using a very limited data set.

Passive sonar array model

Methodology

The proposed localization pipeline consists of three primary stages designed to handle wideband signals and multipath interference under low-snapshot conditions.

1. Chirp Transformation (FRFT)

Sonar systems typically use wideband "chirp" signals where frequency changes over time. The steering vectors for these signals are time-variant, which breaks standard processing tools. The researchers utilize the Fractional Fourier Transform (FRFT) to transform these wideband chirps into single-frequency sine waves in the FRFT domain. This effectively "freezes" the signal into a narrowband-like state with a time-invariant steering vector.

2. Sparse Estimation (SPICE)

Once in the FRFT domain, the SPICE algorithm is applied. SPICE is a sparsity-based technique that doesn't require prior knowledge of the number of targets. Crucially, it provides super-resolution DOA and TOA estimates even with as few as two snapshots, where traditional methods would produce unusable results.

3. Virtual Station (VS) Method

To handle multipath, a "Virtual Station" approach is employed. Instead of trying to filter out reflections, the reflection points are treated as virtual sensors. This geometric trick converts a complex multipath problem back into a simpler LOS problem, allowing for more accurate final coordinate calculation.

Results and Discussion

The study relies on simulated sonar data to validate the algorithm, using a wideband chirp with a 1500 Hz center frequency and an eight-sensor linear array.

In simulations with only two snapshots, the proposed FrFT-SPICE method clearly separated two targets just 2° apart. In contrast, the FrFT-MUSIC algorithm failed to resolve them, merging the two targets into a single, inaccurate peak. The FrFT-SPICE method produced an angular spectrum with significantly lower sidelobes and higher accuracy than MUSIC-based approaches.

Comparison of VS method RMSE and CRLB

Furthermore, the authors calculated the Cramer-Rao Lower Bound (CRLB)—the theoretical limit of estimator accuracy. As shown in the performance analysis, the Virtual Station method's Root Mean Square Error (RMSE) closely tracks the CRLB, proving the method is near-optimal even as multipath conditions vary.

Conclusion

This paper introduces a robust method for underwater sonar localization that achieves high precision using only a handful of signal snapshots. By combining FRFT with the sparsity-based SPICE algorithm and a "virtual station" geometry, the challenges of unstable underwater environments and complex multipath interference are successfully addressed. Future work will focus on real-world validation through sea trials to account for non-Gaussian noise and complex underwater acoustic phenomena like thermoclines.

Summary updated

Abstract

Sonars can be used for detection of targets located on the sea floor where the signal undergoes multipath propagations which may arise large localization errors. Besides, underwater environment is often highly unstable so that a large number of snapshots which means long time sampling are not reliable. In sonar systems, the wideband chirp signals are usually used. This paper presents a nearfield target localization method for underwater environments. First, the time of arrival (TOA) and the direction of arrival (DOA) for nearfield wideband chirp signals are estimated with only a few snapshots by using the fractional Fourier transform (FRFT) and the SPICE algorithm. Then, the virtual station method which is proposed in our previous work is used so that the multipath problems can be converted into line-of-sight (LOS) problem and its Cramer-Rao lower bound (CRLB) is also analyzed. Finally, simulation results demonstrate the effectiveness of the proposed method.
BibTeX
@inproceedings{li2018nearfield,
  author = {Zhuo Li and Deliang Liu and Xiwei Guo and Peng He},
  title = {{Nearfield target localization with a few snapshots for sonar array}},
  booktitle = {EURASIP},
  year = {2018},
  doi = {10.1186/s13638-018-1096-3},
}
Figure