Optimized MA-TRIS resource allocation in 6G networks

Movable antennas (MAs) and transmissive reconfigurable intelligent surfaces (TRISs) have become potentially useful technologies in the sixth-generation wireless systems; however, their joint optimization is underdeveloped because the non-convex coupling occurs between the continuous antenna positioning and the discrete phase-shift configuration. The paper suggests an effective model of the simultaneous optimization of the MA positioning and low-resolution (12 bits) TRIS phase shifts (near-field compact base stations at real propagation, cosine radiation pattern, and imperfect channel state information) under realistic propagation. To eliminate the prohibitively complex grid search that was used in earlier MA-TRIS designs, the authors propose fast continuous optimization algorithms, in particular an improved particle swarm optimization (PSO), compared to which genetic algorithm (GA) and simulated annealing (SA) are benchmarks, that explicitly deal with the mixed continuous–discrete nature of the problem. Long-run simulation has shown that the proposed PSO-based scheme provides near-optimal received signal-to-noise ratio at reduced computational time, thus providing a scalable and viable solution to future near-field intelligent wireless systems.