基于RBF神经网络的认知无线电认知引擎设计

Design of Cognitive Engine Based on Radial Basis Function Neural Network in Cognitive Radio

认知无线电(CR)是一种智能无线通信系统,它能根据环境变化、业务需求动态调整参数,提高系统性能,其核心技术是认知引擎的设计。认知引擎可引入人工智能领域的推理与学习方法来实现CR的感知、自适应与学习能力。为适应变化的无线环境和用户需求,提出基于径向基神经网络(RBF)的CR认知引擎设计方法,该法通过对经验知识和环境的学习,重配置通信参数,以达到资源合理分配,提高系统性能。该引擎由两层RBF神经网络组成,外层神经网络学习全局属性,内层神经网络学习局部属性,以解决路由协议及局部参数的学习配置。在训练RBF神经网络学习模型后,根据定义的两个测试基准函数,评估模型性能,仿真验证了该学习模型的有效性,且能够有效实现CR学习重构。

Cognitive radio （CR） is an intelligent wireless communication system, which can dynamically adjust the parameters to improve system performance depending on the environmental change and quality of service （Qos）. The core technology for CR is the design of cognitive engine, which can introduce reasoning and learning methods to achieve the perception, adaptation and learning. Considering the dynamical environment and demands, a scheme of cognitive engine was proposed based on the radial basis function （RBF） neural network. The scheme could study from experience and environment to reconfigure communication parameters and improve system performance. The cognitive engine was composed of two RBF_NN layers to solve the learning configurations of routing protocol and local parameters. The outer layer learned the global properties, while the inner layer learned the local attributes. After training, the learning model performance was evaluated according to two defined benchmark functions. The simulation results show that the learning model is effective and the cognitive engine can effectively achieve the study and reconfiguration function.