ID 原文 译文
3843 所提框架采用深度强化学习方法实现实时异构切片间资源需求的自动预测与分配，切片内用户资源分配建模为基于形状的二维背包问题并采用启发式算法数值求解，从而实现切片内资源定制化。 In the proposedframework, real-time and automatic inter-slice resource demand prediction and allocation were realized based on deepreinforcement learning (DRL), while intra-slice user resource allocation was modeled as a shape-based 2-dimensionpacking problem and solved with a heuristic numerical algorithm, so that intra-slice resource customization was achieved.
3844 仿真结果表明，基于资源预留的方法能够使 EIoT 切片显式保留资源，提供了更好的安全隔离级别； Simulation results show that the resource reservation-based method enable EIoT slices to explicitly reserveresources, provide a better security isolation level, and DRL could guarantee accuracy and real-time updates of re-source reservations.
3845 深度强化学习能够保证资源预留的准确和实时更新，有效兼顾资源利用率和切片差异化服务质量要求。与 4 个已有算法对比表明，Dueling DQN 具有更好的性能优势。 Compared with four existing algorithms, dueling deep Q-network (DQN) performes better thanthe benchmarks.
3846 为了提高传输的灵活性和应对应急通信的场景，提出了一种基于无人机辅助的非正交多址接入系统用户分组与功率分配联合优化算法。 To improve the transmission flexibility of emergency communication scenarios, a joint user grouping andpower allocation algorithm was proposed for UAV-aided non-orthogonal multiple access (NOMA) system.
3847 考虑基站最大发射功率约束与用户分组约束，构建了 NOMA 多用户和速率最大化的资源分配问题。 A sum-ratemaximization resource allocation algorithm was formulated under the constraints of the maximum transmit power and theuser grouping.
3848 为了解决该非凸问题，基于图论提出了最大割定理下的用户分组算法，实现了组内用户与无人机的相对距离最小。 To solve the non-convex problem, a graph-based user grouping algorithm was proposed to minimize therelative distance between the UAV and users.
3849 基于分组结果，利用辅助变量法，将原非凸功率分配子问题转化为凸优化问题求解。 Based on the grouping result, the power allocation sub-problem was trans-formed into a convex one by using the auxiliary-variable approach.
3850 仿真结果表明，所提算法具有较好的和速率性能。 Simulation results demonstrate that the proposed al-gorithm has better performance in terms of sum rates.
3851 跨域访问控制虽然能提升互操作性，但也可降低域内自治性，因此如何平衡域间互操作性和域内自治性是一个重要的问题。 Cross-domain access control can improves interoperability but reduces intra-domain autonomy.
3852 针对该问题，提出一种基于多目标整数规划优化的跨域访问控制策略映射机制。 To balance in-ter-domain interoperability and intra-domain autonomy, a cross-domain access control policy mapping to the problem ofmulti-objective integer optimization programming was formulated.