认知无线网络中的智能协作频谱感知机制研究

作为智能体，认知无线电应具有智能学习和智能判决的能力。为充分发掘认知用户的智能体特性，本文提出一种基于支持向量机和模糊积分的智能协作频谱感知机制。该机制将协作频谱感知模型转化成基于模糊积分的多分类器融合模型，其中每个认知用户均被看作一个独立的支持向量机分类器，单个感知周期内得到的采样数据作为分类器的输入，分类器的概率输出将被发送至融合中心，融合中心采用模糊积分算法将各分类器得到的结果进行融合并判决。该机制充分挖掘了认知用户在频谱感知阶段的“智能学习”能力和信息融合阶段的“智能判决”能力，仿真结果进一步表明，与单一的分类模型相比，本文提出的智能协作频谱感知机制具有更高的检测概率和更低的虚警概率。 

Research on Intelligent Cooperative Spectrum Sensing Mechanism in Cognitive Radio Networks

As a smart object, cognitive radio should have the ability of intelligent learning and decision-making. To fully exploit the intelligence of the cognitive users, an intelligent cooperative spectrum sensing mechanism is proposed, in which cooperative spectrum sensing model is converted into a multi-classifier fusion model based on fuzzy integral. Every secondary user is viewed as a separate support vector machine classifier. Sampling data obtained during sensing periods will be used as the input of the classifier. And the probability output of the classifier will be sent to the fusion center, where fuzzy integral algorithm is carried to get the final sensing results. The ability of learning in spectrum sensing stage and intelligent decision-making in information fusion stage are fully reflected in this mechanism. And simulation results further show that the proposed mechanism has higher detection probability and lower false alarm probability than a single classifier model.