多模态数据融合年龄感知的节点调度和资源分配

iming at the timeliness of data collection in the data fusion scenario of industrial Internet of things , this paper propose the age of multimodal data fusion (AoMF)to capture the information freshness requirements of a real-time application jointly served by actual collaboration nodes. The validity period of each modal data is distinguished by different upper bounds of age of information (AoI). The problem of minimizing the AoMF is modeled as a Markov decision process (MDP). Under the limitation of communication and computing resources, the scheduling of nodes andnon orthogonal multiple access (NOMA) transmission resource allocation strategy are jointly designed to improve the freshness of fusion tasksbased on deep Q network (DQN) algorithm. Simulation results show that the proposed scheme is effective in reducing AoMF.