Computational electromagnetic simulation of smart antenna systems in urban microcellular environments

The paper presents computational electromagnetics (CEM)-based characterization of smart antenna-system performance in urban microcellular environments. Mutual coupling effects between antenna-array elements are modeled using the method of moments (MoM) and the mobile environment is modeled using electromagnetic ray tracing (ERT). The smart antenna systems simulated in several urban microcells (covering a region of downtown Austin, TX) use uniform circular antenna arrays operating at 1.8 GHz. Direction-of-arrival (DOA) distribution information is used to comment on alternative array geometries for urban microcells. Power distributions are given to illustrate the variability of microcell shape using both DOA- and spatial signaturebased downlink beamforming. In addition, these power distributions demonstrate how multivariate optimization can be used to modify microcell shape and to compensate for the presence of a blocking cellular tower. Spatial signature-variation information is used to characterize the overall environment and to motivate vector-autoregressive (VAR) prediction of mobile usersʹ spatial signatures using a Kalman filter. Results from this developed prediction technique are provided for mobile users in the urban microcellular environment.