为了解决Vehicle to Everything(V2X)毫米波通信系统延时高、链路易阻塞等问题,基于车辆和用户终端周围环境状态信息的感知,提出一种视觉辅助的能效最大阻塞预测方法.利用视觉感知模型实现系统对目标用户以及周围障碍物的精准感知,结合深度强化学习设计了一种融合特征和时间注意力的DA-DBLSTM网络预测未来链路阻塞到达时间,与传统注意力相比,该融合注意力不仅可以关注每个时间单元中的不同特征,而且关注不同时间单元的时序信息,使检测效果更优.仿真和分析结果表明,提出的DA-DBLSTM网络预测链路阻塞效果明显,在均方误差(Mean Square Error,MSE)、均方根误差(Root Mean Square Error,RMSE)、平均绝对误差(Mean Absolute Error,MAE)和平均绝对百分比误差(Mean Absolute Percentage Error,MAPE)方面均优于现有方法.
V2X Communication Blockage Prediction Method Based on Bi-LSTM with Fusion Attention
In order to solve the problems of high delay and easy blocking in Vehicle to Everything(V2X)millimeter wave communication system,a vision assisted blockage prediction method with maximum energy efficiency is proposed based on the perception of the environmental state information around vehicles and user terminals.The visual perception model is used to realize accurate perception of target users and surrounding obstacles,and a DA-DBLSTM network integrating feature attention and time attention is designed with deep reinforcement learning to predict the arrival time of future link blockage.Compared with traditional attention,this fusion attention can not only focus on different features in each time cell,but also focus on the timing information of different time cells,so that the detection effect is better.The simulation and analysis results show that the proposed DA-DBLSTM network can predict link blockage effectively,and is superior to the existing methods in terms of Mean Square Error(MSE),Root Mean Square Error(RMSE),Mean Absolute Error(MAE)and Mean Absolute Percentage Error(MAPE).
万方数据
维普
