References & Further Reading

References

  1. B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Barron, R. Ramamoorthi, and R. Ng, NeRF: Representing scenes as neural radiance fields for view synthesis, 2020

    The original NeRF paper that introduced neural radiance fields with positional encoding and hierarchical volume rendering. The foundation for all RF adaptations in this chapter.

  2. Y. Zhao, X. Zhu, and Y. C. Eldar, NeRF2: Neural radio-frequency radiance fields, 2023

    The foundational paper adapting NeRF for RF propagation modelling from RSS measurements. Discussed in detail in Section 24.3.

  3. T. Mueller, A. Evans, C. Schied, and A. Keller, Instant neural graphics primitives with a multiresolution hash encoding, 2022

    Multi-resolution hash encoding (Instant-NGP) that accelerates NeRF training by orders of magnitude. Applicable to all RF-NeRF variants.

  4. J. T. Barron, B. Mildenhall, M. Tancik, P. Hedman, R. Martin-Brualla, and P. P. Srinivasan, Mip-NeRF: A multiscale representation for anti-aliasing neural radiance fields, 2021

    Anti-aliased volume rendering using conical frustums instead of point samples. Relevant for modelling antenna beam widths in RF applications.

  5. M. Tancik, P. P. Srinivasan, B. Mildenhall, S. Fridovich-Keil, N. Raghavan, U. Singhal, R. Ramamoorthi, J. T. Barron, and R. Ng, Fourier features let networks learn high frequency functions in low dimensional domains, 2020

    Theoretical analysis of why positional encoding enables MLPs to learn high-frequency functions. Explains the spectral bias of neural networks.

  6. T. Orekondy, S. Dhakal, and A. Balan, WiNeRT: Towards neural ray tracing for wireless channel modelling and differentiable simulations, 2023

    Multi-bounce neural ray tracing for wireless propagation, extending NeRF with reflection and diffraction rays.

  7. T. Huang, S. Zhao, Z. Li, and Y. Chen, DART: Doppler-aided radar tomography, 2024

    Neural radiance fields incorporating Doppler information for automotive radar scene reconstruction.

  8. G. Caire, On the illumination and sensing model for RF imaging, 2026

    Unifies diffraction tomography models with MIMO radar models. The Born forward model derived here is connected to the RF volume rendering equation in Section 24.5.

  9. J. W. Goodman, Introduction to Fourier Optics, Roberts and Company, 3rd ed., 2005

    Classical reference for wave propagation, diffraction, and the Fresnel number. Provides the optics background for the optical-to-RF translation in Section 24.2.

  10. J. Zhang, L. Liu, and G. Caire, R-NeRF: Neural radiance fields for RIS-enabled environments, 2024

    Extends RF-NeRF to reconfigurable intelligent surfaces, enabling joint scene reconstruction and RIS optimisation.

  11. S. Sun, H. Xu, and W. Chen, VoxelRF: Voxel-based neural radiance fields for efficient RF scene reconstruction, 2023

    Replaces the MLP with a voxel grid for faster inference, achieving 10x speedup at comparable accuracy.

  12. M. Li, Z. Wang, and Y. Zhao, NeRF-APT: Joint access point localisation and scene reconstruction via neural radiance fields, 2024

    Inverts the NeRF workflow to localise unknown transmitters from RSS measurements at known receiver positions.

  13. Z. Wei, S. Xu, and Y. Chi, Neural implicit SAR: Autofocus and super-resolution from sparse apertures, 2024

    Adapts neural fields for coherent SAR imaging with implicit regularisation and joint autofocus.

  14. N. Rahaman, A. Baratin, D. Arpit, F. Draxler, M. Lin, F. A. Hamprecht, Y. Bengio, and A. Courville, On the spectral bias of neural networks, 2019

    Analysis of the spectral bias of neural networks toward low-frequency functions. Provides the theoretical basis for the implicit regularisation in ISAR-NeRF.

Further Reading

For readers who want to go deeper into specific topics from this chapter.

  • Neural fields survey for 3D vision

    Y. Xie, T. Takikawa, S. Saito, O. Litany, S. Yan, N. Khan, F. Tombari, J. Tompkin, V. Sitzmann, and S. Sridhar, 'Neural fields in visual computing and beyond,' Computer Graphics Forum, vol. 41, 2022

    Comprehensive survey of neural field methods in computer vision, providing context for the RF adaptations in this chapter.

  • RF propagation modelling with deep learning

    S. Bakirtzis, J. Zhang, K. Pollin, and C. Oestges, 'DeepRay: Deep learning meets ray-based radio propagation modelling,' Proc. IEEE VTC, 2022

    Alternative neural RF propagation approach using ray-based features rather than volumetric rendering. Useful contrast with the NeRF-based methods in this chapter.

  • Differentiable rendering for inverse problems

    V. Sitzmann, J. N. P. Martel, A. W. Bergman, D. B. Lindell, and G. Wetzstein, 'Implicit neural representations with periodic activation functions,' Proc. NeurIPS, 2020

    SIREN architecture using periodic activations as an alternative to positional encoding. Relevant for representing high-frequency RF fields.

  • Sionna for ray tracing comparison

    J. Hoydis et al., 'Sionna: An open-source library for next-generation physical layer research,' Proc. IEEE GLOBECOM, 2022

    GPU-accelerated differentiable ray tracer for wireless channels. Provides a baseline for comparing RF-NeRF against physics-based ray tracing.

ExercisesCh 25 →