Part 5: System-Level Design and 6G Architecture
Chapter 26: Massive MIMO Prototyping and Testbeds
Research~240 min
Learning Objectives
- Survey the landscape of massive MIMO testbeds — LuMaMi, ArgosV3, OpenAirInterface, srsRAN, HHI, and the Massive Beams spin-off — and identify which architectural choices drive which capabilities
- Quantify the latency, throughput, and memory budgets imposed by the 5G NR sub-millisecond slot and derive the compute envelope available to channel estimation and precoding
- Derive the SINR loss caused by fixed-point quantization in the MAC pipeline and size the mantissa width needed for a given target rate
- Formulate the TDD reciprocity calibration problem, understand why relative calibration is sufficient, and analyze the residual error versus calibration update period
- Estimate the BER degradation induced by residual timing, frequency, and phase-noise offsets and specify synchronization accuracy requirements for massive MIMO
- Design an over-the-air measurement campaign, curate the raw data, and explain how the resulting dataset feeds the 3GPP Rel-18/19 AI/ML workitems
Sections
Prerequisites
Massive MIMO fundamentals: favorable propagation, channel hardening, TDD reciprocity (MIMO Ch. 1, 3, 4)Linear precoding and uplink detection: MR, ZF, MMSE, their compute cost (MIMO Ch. 6, 9)OFDM and frame structures for multi-carrier wideband signals (MIMO Ch. 10)5G NR MIMO framework: CSI-RS, SRS, frame structure, numerology (MIMO Ch. 22)Cell-free massive MIMO and distributed processing (MIMO Ch. 11, 13, 14)Digital signal processing in fixed-point arithmetic (FSP background)
💬 Discussion
Loading discussions...