Multi-user MIMO dataset collection

A measurement campaign of multi-user MIMO channel sounding and beamforming using the RENEW platform at Rice University.

Four access point radio heads, labeled AP 1 to AP 4, on chairs across an indoor hall at Rice University, connected to a central hub on a cart — The access point (AP) configuration.

Six Iris SDR clients, labeled UE 1 to UE 6, on chairs across the same hall, with a network switch — The user (UE) configuration.

I was collecting extensive datasets for the experimental phase of my current project, “Evaluating Distributed MU-MIMO Beamforming.” We used the RENEW programmable massive MIMO platform: four radio heads, each with a linear array of 16 antennas, connected to a central hub as our access points (APs), and six standalone Iris SDRs as our clients (UEs). We collected uplink (pilot and data) datasets in multiple scenarios:

Centralized: co-located APs with coherent or non-coherent arrays
Distributed: distributed APs with coherent or non-coherent arrays

We made these datasets publicly available for research purposes; they are also on IEEE DataPort as the Distributed Multi-User MIMO Datasets. In our previous work, published at IEEE SAM 2022, we generated extensive datasets that are also available.

Related