Chapter Summary

Chapter Summary

Key Points

• 1.

  The fronthaul link connecting APs to the CPU is the fundamental bottleneck in distributed MIMO. CPRI requires fronthaul rates proportional to $W \cdot N_t$, which exceeds 200 Gbps for a 64-antenna, 100 MHz AP --- motivating compression and functional splits.

• 2.

  Uplink strategies: Quantize-and-forward (QF) directly quantizes the $N_t$-dimensional observation; estimate-and-forward (EF) first applies local MMSE combining to reduce the dimension to $K$ before quantization. EF is preferred in the massive MIMO regime ($N_t \gg K$). Wyner-Ziv compression further reduces fronthaul by exploiting inter-AP correlation.

• 3.

  Downlink strategies: Compression-based precoding has the CPU compute precoded signals, compress them, and forward to the APs. Finite fronthaul wastes a fraction $2^{-C_{\text{fh}}}$ of each AP's power as compression noise. Joint precoding-compression optimization via alternating methods is the practical approach.

• 4.

  Load balancing: The Goettsch/Li/Caire framework jointly optimizes fronthaul allocation and computation resources across APs. Waterfilling fronthaul capacity across APs provides 15--30% sum rate gains over uniform allocation in heterogeneous networks.

• 5.

  Open RAN: The O-RAN 7.2x functional split places FFT/CP at the RU and MIMO processing at the DU, providing a standardized architecture for cell-free deployments. The split choice directly determines the fronthaul-computation tradeoff.