Dynamic Ego-Velocity estimation Using Moving mmWave Radar: A Phase-Based Approach

TL;DR

The paper tackles robust ego-motion estimation for mobile platforms without relying on visual or inertial sensors. It introduces mmPhase, a phase-based velocity estimation method operating on single-chip mmWave radar data, computing velocity from phase changes via $\frac{d\phi}{dt} = \frac{4\pi v_b}{\lambda}$ and the relation $\phi = \frac{4\pi d}{\lambda}$ with a granularity of $86~\mu s$. The authors implement a real-time prototype and demonstrate that mmPhase reduces velocity estimation error by approximately $4\times$ compared to a Doppler-based baseline and generally outperforms IMU and milliEgo baselines in various scenarios. This work highlights the potential of lightweight mmWave sensing for robust, low-latency ego-velocity estimation in robotics and wearables, and points to future extensions such as multi-object scenarios and physics-informed neural approaches.

Abstract

Precise ego-motion measurement is crucial for various applications, including robotics, augmented reality, and autonomous navigation. In this poster, we propose mmPhase, an odometry framework based on single-chip millimetre-wave (mmWave) radar for robust ego-motion estimation in mobile platforms without requiring additional modalities like the visual, wheel, or inertial odometry. mmPhase leverages a phase-based velocity estimation approach to overcome the limitations of conventional doppler resolution. For real-world evaluations of mmPhase we have developed an ego-vehicle prototype. Compared to the state-of-the-art baselines, mmPhase shows superior performance in ego-velocity estimation.

Figures (3)

• Figure 1: System overview of mmPhase
• Figure 2: (a) mmPhase hardware setup, (b) phase variation with the number of frames.
• Figure 3: (a) Mean Absolute Error in velocity estimation with respect to baselines, (b) Variation in the estimated velocity w.r.t. doppler-based approach at lower velocities.