ID 原文 译文
16955 仿真结果表明所提多小区混合NOMA网络资源分配方案能有效提升全网用户服务体验和公平性。 Simulation results demonstrate that the proposed algorithm can effectivelyimprove the system performance and fairness.
16956 针对5G通信技术高传输速率、多业务场景的挑战，该文提出一种组件化的软件定义无线接入网络新架构。 In view of the challenges of 5G communication technology with high speed and multiple servicescenarios, this paper proposes a new component-based Software Defined wireless access Network (SDN)architecture.
16957 该架构在5G接入网集中单元(CU)，分布单元(DU)，有源天线单元(AAU)架构的基础上，进一步朝组件化方向演进，形成一种由集中控制单元(CCU), CU, DU，射频单元(RU)，AAU等组件化通信单元组成的新架构。 Based on the architecture of Centralized Unit (CU), Distributed Unit (DU) and Active AntennaUnit(AAU) in 5G access network, further component-based evolution is carried out to form a new architecturecomposed the communication units of Centralized Control Unit (CCU), CU, DU, Radio Unit(RU), and AAU.
16958 这种新架构既有利于切片化、虚拟化实现无线接入网，又有利于应用分布式计算技术和硬件加速技术突破通用处理器的计算能力瓶颈，还能降低DU与AAU之间的前传压力。 This new architecture is not only conducive to the realization of wireless access network with slicing andvirtualization, but also conducive to the adoption of distributed computing technology and hardwareaccelerating technology to break through the processing bottleneck of general-purpose processor, and reduce theforward transmission pressure between DU and AAU.
16959 该文还研制了基于此架构的组件化软基站试验原型并进行了测试， In this paper, a prototype of component-based soft basestation is developed and tested.
16960 结果表明该组件化方案在提供高度灵活性的同时，还能够提升通用处理器软基站的吞吐能力，并有效降低远端站址传输流量。 The results show that the component-based scheme can not only provide highflexibility, but also improve the processing capacity of the general processor soft base station and reduceeffectively the traffic of remote stations.
16961 现有的基于分割的场景文本检测方法仍较难区分相邻文本区域，同时网络得到分割图后后处理阶段步骤复杂导致模型检测效率较低。 Existing segmentation based methods have problems, such as the difficulty in distinguishing adjacenttext areas and the low efficiency of model detection caused by the complex steps in the post-processing stage.
16962 为了解决此问题，该文提出一种新颖的基于全卷积网络的场景文本检测模型。 Inorder to solve this problem, this article proposes a novel scene text detection model based on fully convolutionalnetwork, which can solve the problem that adjacent texts are difficult to distinguish in existing methods andimprove the detection speed of the model.
16963 首先，该文构造特征提取器对输入图像提取多尺度特征图。 First, it constructs a feature extractor to extract multi-scale featuremap from the input image.
16964 其次，使用双向特征融合模块融合两个平行分支特征的语义信息并促进两个分支共同优化。 Secondly, the bidirectional feature fusion module is used to fuse the semanticinformation of the two parallel branches and promote the joint optimization of the two branches.