发表刊物:IEEE COMMUNICATIONS LETTERS
摘要:To solve the policy optimizing problem in many scenarios of smart wireless network management using a single universal algorithm, this letter proposes a universal learning framework, which is called AI framework based on deep reinforcement learning (DRL). This framework can also solve the problem that the state is painful to design in traditional RL. This AI framework adopts convolutional neural network and recurrent neural network to model the potential spatial features (i.e., location information) and sequential features from the raw wireless signal automatically. These features can be taken
论文类型:文章
第一作者:Gang Cao
文献类型:J
卷号:22
期号:2
页面范围:400-403
ISSN号:1089-7798
是否译文:否
收录刊物:SCI