ID 原文 译文
21775 现有硬件任务调度算法任务描述不完善且忽视时间维上紧凑性。 The existing hardware task scheduling algorithms describe task imperfectly and ignore the compactness of time dimension.
21776 该文考虑任务下载时间、完善任务属性，以器件2维资源与时间建立3维资源模型，将任务布局问题抽象成特殊的3维空间放置问题，在此模型上分析出现有算法不能克服任务不可预知性和资源占用多变性，导致调度成功率和资源利用率低。 The task downloading time is considered for improving the task attribute, and the 3D-resource model with the two dimensional resource of device and time is established, in order to abstract the issue of task layout into a special three-dimensional space placement issue. With this model, it is concluded that the existing algorithms can not overcome the unpredictability of the task and the diversity of resource occupancy, leading low scheduling success rate and resource utilization rate.
21777 针对此问题，该文提出了一种3维可重构任务调度算法3D_RTSA。设计并实现了基于任务紧迫度的调度策略和基于3D碎裂度的布局策略。 To solve the problem, a three dimensional reconfigurable task scheduling algorithm called 3D_RTSA is proposed. A scheduling strategy based on task urgency and a layout strategy based on 3D fragmentation are designed and implemented.
21778 与其他4种算法实验对比结果表明，在重负载、小任务C30情况下，3D_RTSA调度成功率比GC, Look-aheadest, SPSA, DTI算法分别高3%, 21%, 28%, 35%左右； Compared with the other 4 algorithms, the results show that the scheduling success rate of 3D_RTSA is 3%,21%, 28%, 35% higher than that of GC, Look-aheadest, SPSA and DTI algorithms under the condition of heavy load and small task C30, and the utilization ratio of resources is 5% and 18% higher than that of Look-aheadest and SPSA algorithm under the condition of light load and large task C50.
21779 在轻负载、大任务C50情况下，资源利用率比Look-aheadest, SPSA算法分别高5%, 18%左右，且该文算法时间复杂度并未增加。 Besides, the time complexity of the algorithm is not increased.
21780 为满足次用户不同时延需求并提高信道利用率，该文提出一种基于次用户分级的 PRP M/G/m 排队论频谱切换模型。 In order to meet different delay requirements of secondary users and improve the channel utilization, a PRP M/G/m queuing theory spectrum handoff model based on classified secondary user is proposed.
21781 首先，根据次用户主要业务的时延需求不同，对次用户进行分级评价，将次用户分为时延敏感型和非时延敏感型用户，时延敏感型次用户具有更高的优先级来接入信道。 Firstly, according to different delay requirements of the main service, secondary users are classified into delay-sensitive and non-delay-sensitive users. The delay-sensitive users have higher priority to access the channel.
21782 其次，采用 PRP M/G/m 排队论来建立基于用户分级的频谱切换模型。 Secondly, PRP M/G/m queuing theory is adopted to establish the spectrum handoff model.
21783 同时推导和分析了该模型下不同优先级次用户扩展数据传输时间，进而提出了面向各级次用户自适应频谱切换策略。 Meanwhile, the extended data delivery time of secondary users with different priorities is deduced and analyzed. Then the adaptive spectrum handoff strategy for secondary users is proposed.
21784 仿真结果表明，相较于已有频谱切换策略，该模型能有效降低次用户切换时延以及扩展数据传输时间，保证高优先级次用户的服务质量，提升次用户频谱切换的综合性能。 Simulation results suggest that the proposed model can reduce the handoff delay and extended data delivery time, which can effectively ensure the QoS of high-priority secondary users and improve the overall performance.