| Popis: |
Modern evaluation methodologies for system level simulations of mobile radio networks (e.g., [1]) support modeling systems with multiple simultaneously operating base stations and mobile devices where accurate simulation of multi-user MIMO and other space-time signal processing techniques should be provided. To be able to predict realistic performance, such models introduce a number of channel parameters with complex auto- and cross-correlation properties. Particularly, such models include a set of large-scale parameters (tap delay spread, angular spreads of rays at the base station and mobile station sides, shadow fading, etc.) being cross-correlated as well as auto-correlated in the spatial domain (across the deployment area) with different correlation functions. This paper is devoted to a method for accurate generation of correlated channel parameters during a simulation based on filtering of the parameters defined in the nodes of a spatial grid. The proposed method combines formal mathematical exactness at least in the asymptotic case and a much lower computational complexity in comparison with existing solutions [2]. Method description, analytical derivation, implementation details and verification by numerical simulations for an exemplary deployment scenario are provided in the paper. The described method may be recommended for application in system level simulators for 3G and 4G mobile radio access networks such as WCDMA HSPA, IEEE 802.16m, LTE-A. |
