Abstract
Reconfigurable intelligent surface (RIS) has recently
drawn substantial attention toward performance enhancement
in wireless communication systems. One of the key challenges in
RIS-assisted systems is to design efficient joint beamforming algorithms. However, most existing algorithms rely on instantaneous
channel state information (CSI) and iterative optimization methods, which suffer from high computational complexity. Therefore,
by exploiting the property of the recent popular fluid antennas,
this paper proposes a low-complexity joint beamforming scheme
for an RIS-assisted fluid antenna system (RIS-FAS) requiring
only statistical CSI. Specifically, by carefully adjusting the FA
array at the transmitter side, it is possible to respectively
steer its main-lobe (ML) and grating-lobe (GL) towards the
user and RIS, where the line-of-sight (LoS) channels from the
transmitter to the user and RIS are guaranteed to be identical, allowing independent beamforming for the three cascade
channels. Theoretical analysis and simulation results validate
that the proposed scheme achieves both high-performance and
low-complexity characteristics compared with its conventional
counterparts.