Prerequisites & Notation

Before You Begin

This chapter analyzes the fundamental bottleneck of FDD massive MIMO — the overhead of downlink pilot transmission and uplink CSI feedback — and surveys the principal solutions: compressed feedback, codebook-based reporting (5G NR), deep learning compression, and JSDM-based dimensionality reduction. A solid understanding of the JSDM framework from Chapter 7 and the TDD/FDD contrast from Chapter 1 is essential.

Massive MIMO system model: $\mathbf{H} \in \mathbb{C}^{N_r \times N_t}$ , channel hardening, favorable propagation(Review ch01)
Self-check: Can you state the channel hardening law $\frac{1}{N_t} \mathbf{H}_{k}^{H} \mathbf{H}_{k} \to 1$ and explain its implication for CSI acquisition?
TDD reciprocity and the FDD overhead barrier(Review s05)
Self-check: Can you explain why TDD avoids the $N_t$ -scaling overhead that FDD suffers?
Angular-domain channel representation and spatial covariance structure(Review ch02)
Self-check: Can you write $\mathbf{H}_{k} = \mathbf{U}_k \mathbf{\Lambda}_k^{1/2} \mathbf{G}_k$ and explain the role of each factor?
JSDM: two-stage precoding, group structure, pre-beamforming matrix $\mathbf{B}_g$ (Review ch07)
Self-check: Can you explain how JSDM reduces the effective channel dimension from $N_t$ to $r_g$ ?
Linear precoding: MRT, ZF, MMSE(Review ch06)
Self-check: Can you write the ZF precoder $\mathbf{W} = \mathbf{H}(\mathbf{H}^{H} \mathbf{H})^{-1}$ and state when it is optimal?
Rate-distortion theory: $R(D)$ function, achievability, converse
Self-check: Can you state the rate-distortion function for a Gaussian source and explain its operational meaning?

•

Massive MIMO fundamentals: scaling laws, channel hardening, favorable propagation, achievable rates

•

JSDM framework: group-based precoding, spatial covariance exploitation

•

Rate-distortion theory provides the information-theoretic framework for CSI compression

Notation for This Chapter

Symbols introduced in this chapter. See also the NGlobal Notation Table master table in the front matter.

Symbol	Meaning	Introduced
$\tau_d$	Downlink pilot overhead (number of DL training symbols)	s01
$B_{\text{fb}}$	Number of uplink feedback bits per coherence block	s01
$T_c$	Coherence interval length (in symbols)	s01
$\hat{\mathbf{H}}$	CSI estimate at the base station (reconstructed from feedback)	s01
$\mathbf{H}_{a}$	Angular-domain channel: $\mathbf{H}_{a} = \mathbf{F}^H \mathbf{H}$ where $\mathbf{F}$ is the DFT matrix	s02
$\\mathbf{F}$	Unitary DFT matrix ( $N_t \times N_t$ )	s02
$\\boldsymbol{\\Phi}$	Compression (measurement) matrix for CSI feedback	s02
$\\mathcal{C}$	Codebook: finite set of candidate beamforming vectors	s03
$d_c(\\mathbf{H}, \\mathcal{C})$	Chordal distance between channel subspace and nearest codebook entry	s03
$\\mathcal{E}(\\cdot)$	CsiNet encoder (UE-side neural network)	s04
$\\mathcal{D}(\\cdot)$	CsiNet decoder (BS-side neural network)	s04
$\\gamma$	Compression ratio $\gamma = M / (2 N_t)$ for CsiNet	s04
$\\mathbf{B}_g$	Pre-beamforming matrix for group $g$ (from JSDM)	s05
$r_g$	Effective rank (number of dominant eigenmodes) of group $g$ 's covariance	s05
$\\mathbf{H}_{\\text{eff},k}$	Effective reduced-dimension channel: $\mathbf{H}_{\text{eff},k} = \mathbf{H}_{k} \mathbf{B}_g$	s05

← Ch 7 The FDD Challenge