I will talk about different sources of phase errors in distributed MIMO: channel aging, reciprocity errors, and LO phase drifts.
Specifically I will explain why reciprocity-based beamforming in D-MIMO requires the antenna panels to be phase-aligned, and how this can be accomplished using over-the-air measurements between panels.
I will discuss techniques for such over-the-air alignment, scalabilty aspects of the problem for large D-MIMO systems (massive synchrony), and how this problem can be analyzed using tools from spectral graph theory. I will also describe some common misconceptions around phase alignment problems. Finally, I will comment on the role of phase coherency for a “sibling application”: over-the-air computing for edge intelligence, relying on transmitter channel inversion for uplink data to combine phase-coherently at a base station receiver.
I will talk about different sources of phase errors in distributed MIMO: channel aging, reciprocity errors, and LO phase drifts.
Specifically I will explain why reciprocity-based beamforming in D-MIMO requires the antenna panels to be phase-aligned, and how this can be accomplished using over-the-air measurements between panels.
I will discuss techniques for such over-the-air alignment, scalabilty aspects of the problem for large D-MIMO systems (massive synchrony), and how this problem can be analyzed using tools from spectral graph theory. I will also describe some common misconceptions around phase alignment problems. Finally, I will comment on the role of phase coherency for a “sibling application”: over-the-air computing for edge intelligence, relying on transmitter channel inversion for uplink data to combine phase-coherently at a base station receiver.