Prerequisites & Notation

This chapter surveys the open problems and future directions of RF imaging. It draws on material from the entire book.

Hardware, datasets, simulation, and evaluation (Chapter 31) (Review ch31)
Self-check: Can you explain the inverse crime and list three evaluation metrics for RF image quality?
Deep unfolding and learned OAMP (Chapter 18) (Review ch18)
Self-check: Can you describe how unrolled OAMP incorporates learned denoisers?
Neural scene representations: NeRF, implicit geometry, 3DGS (Chapters 24-26) (Review ch24)
Self-check: Can you explain how RF-NeRF represents a scene as a continuous neural field?
ISAC waveform design and beamforming (Chapter 29) (Review ch29)
Self-check: Can you formulate the communication-sensing trade-off as a Pareto problem?

This chapter is primarily a survey of open problems and introduces minimal new notation. It draws on notation from all preceding chapters.

Symbol	Meaning	Introduced
$\mathcal{D}_{\mathrm{sim}}$	Simulated training data distribution	s01
$\mathcal{D}_{\mathrm{real}}$	Real-world data distribution	s01
$\Delta_{\mathrm{s2r}}$	Sim-to-real performance gap	s01
$\boldsymbol{\gamma}(\mathbf{x}, t)$	Time-varying scene reflectivity	s02
$\{\mathcal{P}_k\}_{k=1}^K$	Set of geometric primitives (boxes, cylinders, planes)	s03
$\mathcal{I}(\mathbf{A}, \boldsymbol{\gamma})$	Imaging Fisher information	s06