DOA estimation with synthetic nested array using moving sensor

Zhan Shi, Xiaofei Zhang, You Sun2017 · DOI

Abstract

Synthetic aperture methods are widely used to extend array aperture by exploiting platform motion. This paper proposes a synthetic nested array using a moving sensor for Direction of Arrival (DOA) estimation. By carefully timing signal samples as a single sensor moves, a longer nested array is constructed that is equivalent to a physical array. This allows the application of traditional DOA estimation algorithms, such as spatial smooth MUSIC (SS-MUSIC). Performance analysis and simulations demonstrate that the synthetic nested array achieves performance parity with real physical arrays, significantly improving resolution and degrees of freedom compared to standard uniform linear arrays.

Introduction

In array signal processing, the resolution of DOA estimation is fundamentally limited by the physical aperture of the sensor array. High-resolution requirements typically necessitate larger arrays, which impose significant constraints on size, weight, and power (SWaP) for mobile platforms like UAVs or satellites. While virtual array expansion techniques can enhance information extraction, they do not increase the physical baseline. This research addresses the limitation by exploiting platform motion to synthesize a nested array aperture from a single moving sensor.

Methodology

The proposed approach integrates the concepts of synthetic aperture and nested arrays to achieve high-resolution DOA estimation with minimal hardware.

Moving Sensor Model

Consider a single sensor moving at a constant velocity $v$ along the X-axis. As the sensor moves, it samples the incoming signal $S(t)$ at specific time intervals. The spatial position of the sensor at any given time $t$ is $x(t) = vt$.

Geometry of the moving sensor and synthetic array construction

Synthetic Nested Array Construction

To emulate a nested array, the sampling is organized into segments. A nested array consists of two concatenated uniform linear arrays (ULAs): an "inner" array with dense spacing and an "outer" array with sparse spacing. By setting the sampling period $T$ such that the distance traveled $vT$ equals the desired unit sensor spacing $d$, the system records data at positions corresponding to a physical nested array.

Structure of a nested array with inner and outer sub-arrays

The data matrix is constructed by taking $L$ snapshots at each of the $M$ virtual sensor positions. This synthetic data is then processed using the Spatial Smooth MUSIC (SS-MUSIC) algorithm, which is particularly effective for nested arrays as it exploits the increased degrees of freedom provided by the difference co-array.

Performance Analysis and Results

The effectiveness of the synthetic nested array was evaluated through numerical simulations.

Simulation Setup

The study simulated 9 far-field narrowband targets distributed between -40° and 40°. The carrier frequency was set to 1 GHz, with a sensor velocity of 100 m/s and a sampling period of 1.5 ms. The synthetic array performance was compared against a physical nested array and a synthetic Uniform Linear Array (ULA).

Key Findings

  • Resolution and Degrees of Freedom: The synthetic nested array successfully resolved all 9 targets using only 6 virtual sensor positions. In contrast, a standard ULA with the same number of sensors would be limited to detecting fewer sources.
  • Performance Parity: Mathematical analysis of the Cramer-Rao Bound (CRB) and simulation results for Root Mean Square Error (RMSE) show that the synthetic array performs identically to a real physical array under ideal conditions.
  • Aperture Extension: By utilizing the physical displacement of the sensor over time, the method effectively increases the baseline, leading to higher angular resolution without increasing hardware complexity.

Discussion and Limitations

While the synthetic nested array offers significant advantages in hardware reduction, it introduces dependencies on temporal and motion stability.

  • Temporal Coherence: The signal must remain stable (coherent) throughout the entire synthesis period. Rapidly changing signals may degrade the estimation accuracy.
  • Motion Accuracy: The model assumes a perfectly known, constant velocity. Real-world factors such as platform vibration or velocity fluctuations could introduce model errors.
  • Signal Bandwidth: The current implementation is optimized for narrowband signals. Extending this to wideband signals remains a subject for future research.

Conclusion

The proposed synthetic nested array provides a robust framework for high-resolution DOA estimation using a single moving sensor. By trading processing time and platform mobility for hardware complexity, this method enables advanced sensing capabilities on small, mobile platforms.

Summary updated

Abstract

The synthetic aperture methods are widely used to extend array aperture by exploiting platform motion. In this paper, a synthetic nested array using moving sensor for DOA estimation is proposed. With the moving sensor, we can construct a longer nested array which is equivalent to the real array. Then traditional DOA estimation algorithms for nested array can also be applied, such as spatial smooth MUSIC algorithm. By extending the array aperture and exploiting synthetic nested array, the DOA estimation performance can be improved. In the ideal case, the performance of synthetic and real array is the same which is vilified by performance analysis and simulations.
BibTeX
@misc{shi2017doa,
  author = {Zhan Shi and Xiaofei Zhang and You Sun},
  title = {{DOA estimation with synthetic nested array using moving sensor}},
  year = {2017},
  doi = {10.2991/mecs-17.2017.139},
}
Figure