Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

HYPERDOA: Robust and Efficient DoA Estimation using Hyperdimensional Computing

Rajat Bhattacharjya, Woohyeok Park, Arnab Sarkar, Hyunwoo Oh, Mohsen Imani, Nikil Dutt

TL;DR

This work tackles robust Direction of Arrival estimation under low-SNR and coherent-source conditions, where classical subspace methods falter and neural approaches are energy-hungry and opaque. It introduces HYPERDOA, a Hyperdimensional Computing–based DoA estimator that reframes the problem as pattern recognition and employs two feature-extraction variants to capture spatial information. The approach uses Fourier Holographic Reduced Representations for encoding, an associative memory of angle centroids, and a greedy multi-source decoding stage, providing a transparent and compute-efficient inference pipeline. Demonstrations on synthetic ULAs show improved accuracy in challenging scenarios and substantial energy savings on an embedded Jetson Xavier NX, underscoring its suitability for edge and safety-critical applications.

Abstract

Direction of Arrival (DoA) estimation techniques face a critical trade-off, as classical methods often lack accuracy in challenging, low signal-to-noise ratio (SNR) conditions, while modern deep learning approaches are too energy-intensive and opaque for resource-constrained, safety-critical systems. We introduce HYPERDOA, a novel estimator leveraging Hyperdimensional Computing (HDC). The framework introduces two distinct feature extraction strategies -- Mean Spatial-Lag Autocorrelation and Spatial Smoothing -- for its HDC pipeline, and then reframes DoA estimation as a pattern recognition problem. This approach leverages HDC's inherent robustness to noise and its transparent algebraic operations to bypass the expensive matrix decompositions and ``black-box'' nature of classical and deep learning methods, respectively. Our evaluation demonstrates that HYPERDOA achieves ~35.39% higher accuracy than state-of-the-art methods in low-SNR, coherent-source scenarios. Crucially, it also consumes ~93% less energy than competing neural baselines on an embedded NVIDIA Jetson Xavier NX platform. This dual advantage in accuracy and efficiency establishes HYPERDOA as a robust and viable solution for mission-critical applications on edge devices.

HYPERDOA: Robust and Efficient DoA Estimation using Hyperdimensional Computing

TL;DR

This work tackles robust Direction of Arrival estimation under low-SNR and coherent-source conditions, where classical subspace methods falter and neural approaches are energy-hungry and opaque. It introduces HYPERDOA, a Hyperdimensional Computing–based DoA estimator that reframes the problem as pattern recognition and employs two feature-extraction variants to capture spatial information. The approach uses Fourier Holographic Reduced Representations for encoding, an associative memory of angle centroids, and a greedy multi-source decoding stage, providing a transparent and compute-efficient inference pipeline. Demonstrations on synthetic ULAs show improved accuracy in challenging scenarios and substantial energy savings on an embedded Jetson Xavier NX, underscoring its suitability for edge and safety-critical applications.

Abstract

Direction of Arrival (DoA) estimation techniques face a critical trade-off, as classical methods often lack accuracy in challenging, low signal-to-noise ratio (SNR) conditions, while modern deep learning approaches are too energy-intensive and opaque for resource-constrained, safety-critical systems. We introduce HYPERDOA, a novel estimator leveraging Hyperdimensional Computing (HDC). The framework introduces two distinct feature extraction strategies -- Mean Spatial-Lag Autocorrelation and Spatial Smoothing -- for its HDC pipeline, and then reframes DoA estimation as a pattern recognition problem. This approach leverages HDC's inherent robustness to noise and its transparent algebraic operations to bypass the expensive matrix decompositions and ``black-box'' nature of classical and deep learning methods, respectively. Our evaluation demonstrates that HYPERDOA achieves ~35.39% higher accuracy than state-of-the-art methods in low-SNR, coherent-source scenarios. Crucially, it also consumes ~93% less energy than competing neural baselines on an embedded NVIDIA Jetson Xavier NX platform. This dual advantage in accuracy and efficiency establishes HYPERDOA as a robust and viable solution for mission-critical applications on edge devices.

Paper Structure

This paper contains 9 sections, 4 equations, 3 figures.

Table of Contents

  1. Introduction
  2. System Model
  3. HYPERDOA Architecture
  4. Feature Extraction
  5. HDC Encoding
  6. Associative Memory
  7. Multi-Source Decoding
  8. Experiments and Evaluation
  9. Conclusion and Future Work

Figures (3)

  • Figure 1: The HYPERDOA Architecture. A raw signal snapshot matrix $\mathbf{X}$ is first processed by the Feature Extraction module to generate a low-dimensional feature vector $\mathbf{f}$. This vector is then mapped by the HDC Encoder into a high-dimensional hypervector $\mathcal{H}_q$. The Associative Memory compares $\mathcal{H}_q$ against its stored prototype hypervectors (centroids) to produce an angular pseudo-spectrum of similarity scores. Finally, the Multi-Source Decoding stage identifies peaks in this spectrum to estimate the Direction of Arrival (DoA) for multiple sources.
  • Figure 2: DoA Estimation MSPE ($T=100$, $N=8$, coherent sources) for SNR= [1,5] dB (upper row) and [-5,-1] dB (lower row).
  • Figure 3: Accuracy vs. Energy Comparison ($M$=4, SNR=[-5,5] dB, $T$=100, $N$=8, coherent sources).