Strategies to accelerate MIMO channel capacity optimization on GPUs are outlined. The optimization scheme is dealt with by properly facing the main computational issues. In particular, the propagation environment is described by ultrafast Geometrical Optics (GO), singular values are computed by a very fast scheme and the optimizer is a parallel version of the differential evolutionary algorithm. The unknowns are given proper representations to reduce the number of optimization parameters.
GPU-based electromagnetic optimization of MIMO channels / Breglia, A.; Capozzoli, A.; Curcio, C.; Liseno, A.; Di, Donna. - In: APPLIED COMPUTATIONAL ELECTROMAGNETICS SOCIETY JOURNAL. - ISSN 1054-4887. - 33:2(2018), pp. 172-175.
GPU-based electromagnetic optimization of MIMO channels
capozzoli a.
;curcio c.;liseno a.;
2018
Abstract
Strategies to accelerate MIMO channel capacity optimization on GPUs are outlined. The optimization scheme is dealt with by properly facing the main computational issues. In particular, the propagation environment is described by ultrafast Geometrical Optics (GO), singular values are computed by a very fast scheme and the optimizer is a parallel version of the differential evolutionary algorithm. The unknowns are given proper representations to reduce the number of optimization parameters.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
