多径衰落非线性动力特征初探

Research on nonlinear dynamical characteristics of multipath fading channels

本文应用现场实测数据初步探讨了蜂窝移动通信环境中多径衰落的非线性动力特征。首先通过重构状态空间和分析关联维数验证了多径衰落动力机制具有有限维自由度,然后设计了前向BP神经网络来拟合其非线性动力系统,并重构了多径信号。分析和仿真结果表明,与传统的随机模型相比,多径衰落的内在物理机制更多地表现出非线性系统特征。

This paper investigates the nonlinear dynamical system characteristics of mulitpath fading channels based on the measured data from field trials. By reconstructing the status space and analyzing correlation dimension, we show that its dynamical mechanism has finite degree of freedom. Then a BP neural network is designed to approximate its nonlinear dynamical system and multipath signals are reconstructed, yielding reasonably accurate replications. The analysis and simulation results suggest that the nonlinear dynamical system could be a more suitable model than the conventional random process model for describing the multipath fading phenomenon.