基于多头自注意力机制的协作频谱感知算法  

Cooperative spectrum sensing algorithm based on multi-head self-attention mechanism

作　　者：李涛 张煜培[1] 赵知劲[1,2] Li Tao;Zhang Yupei;Zhao Zhijin(School of Communication Engineering,Hangzhou Dianzi University,Hangzhou 310018,China;National Key Laboratory of Communication System Information Control Technology,36th Research Institute of China Electronics Technology Group,Jiaxing 314001,China)

机构地区：[1]杭州电子科技大学通信工程学院,浙江杭州310018 [2]中国电子科技集团第36研究所通信系统信息控制技术国家级重点实验室,浙江嘉兴314001

出　　处：《电子技术应用》2023年第11期88-93,共6页Application of Electronic Technique

基　　金：国家自然科学基金项目(U19B2016)。

摘　　要：针对协作频谱感知中部分用户受到阴影、衰落等影响导致在低信噪比环境下对微弱信号的感知性能急剧降低及部分聚类模型没有充分利用能量向量的问题,提出一种基于多头自注意力机制的协作频谱感知算法。各次用户独立采样数据,融合中心收集各次用户的能量数值并构成能量向量作为自注意力网络的输入;设计基于多头自注意力机制的网络模型,利用该网络自动学习信号和噪声的能量向量特征,有效提取局部特征,实现智能协作频谱感知。仿真结果表明,该频谱感知算法性能优于对比算法,在信噪比为-14 dB、虚警概率为0.001时,该算法检测概率高于对比算法0.29~0.4。Aiming at the problem that some users are affected by shadows,fading and other factors in cooperative spectrum sens‐ing,the sensing performance of weak signals in low signal-to-noise ratio environment is dramatically reduced,and some cluster‐ing models do not make full use of energy vectors,a cooperative spectrum sensing algorithm based on multi-head self-attention mechanism is proposed.Each user samples data independently,and the fusion center collects the energy value of each user and forms the energy vector as the input of the self-attention network.A network model based on multi-head self-attention mechanism is designed.The network is used to automatically learn the energy vector features of signals and noises,effectively extract local features,and realize intelligent cooperative spectrum sensing.The simulation results show that the performance of the spectrum sensing algorithm is better than that of the comparison algorithm.When the SNR is-14 dB and the false alarm probability is 0.001,the detection probability of the algorithm is 0.29 to 0.4 higher than that of the comparison algorithm.

关 键 词：认知无线电网络 多头自注意力机制 协作频谱感知 能量向量 

分 类 号：TN925[电子电信—通信与信息系统]