An autonomic joint radio resource management algorithm in end-to-end reconfigurable system

This paper presents the multi-step Q-learning （MQL） algorithm as an autonomic approach to the joint radio resource management （JRRM） among heterogeneous radio access technologies （RATs） in the B3G environment. Through the ＂trial-and-error＂ on-line learning process, the JRRM controller can converge to the optimized admission control policy. The JRRM controller learns to give the best allocation for each session in terms of both the access RAT and the service bandwidth. Simulation results show that the proposed algorithm realizes the autonomy of JRRM and achieves well trade-off between the spectrum utility and the blocking probability comparing to the load-balancing algorithm and the utility-maximizing algorithm. Besides, the proposed algorithm has better online performances and convergence speed than the one-step Q-learning （QL） algorithm. Therefore, the user statisfaction degree could be improved also.     

基于Swin Transformer和CNN的汉字书法教学系统

针对日益增长的汉字书法学习需求,将滑动窗口自注意力(Swin Transformer, ST)模型和卷积神经网络(Convolutional Neural Network, CNN)模型相结合,提出手写体汉字识别ST-CNN模型,进而开发了汉字书法教学系统。实测结果表明,ST-CNN模型识别准确率约为91.6%,较传统的ST模型提升了约0.5个百分点,较传统的CNN模型与ST模型,在收敛速度上分别提升了约10和30个百分点,开发的汉字书法教学系统性能良好。