Search for: All records

In MIMO communications, it is expensive and sometimes impossible to obtain timely channel state information at transmitter (CSIT) when the number of antennas is large. Furthermore, if the carrier frequency is high, the variation of the phases of the channel state information is significant with small movement in the channels. On the other hand, spatial signal path directions and the amplitudes of path gains vary slowly. We propose a beam network methodology for multi-antenna communication using such partial CSIT as directions and amplitudes instead of the conventional approach that employs full CSIT. Three representative problems are presented to demonstrate the simplicity and performance of the methodology. These include the single-user MIMO channel, a MIMO interference channel, and a multicast network via a reconfigurable intelligent surface (RIS). In particular, we see that the usual high complexity of RIS state design can be simplified to choosing states to connect an incoming beam to outgoing beams.