ID 原文 译文
25095 为了解决数据集样本不够的问题采用循环生成对抗网络 (Cycle Generative Adversarial Networks，Cy-cleGAN)算法对 GPＲ B-SCAN 图像数据进行增强。 In order to solve the problem of insufficient samples in the dataset, the GPＲ B-SCAN image data is augmented by using cycle generative adversarial networks algorithm (CycleGAN)．
25096 采用 Faster Ｒ-CNN 算子来定位双曲线图像区域，充分利用双曲线结构对称性及其方向差异性特征，设计与之对应的卷积核模板，通过卷积运算实现对 B-SCAN 图像中双曲线目标的有效分割。 The Faster Ｒ-CNN operator is used to locate the hyperbolic image area, making full use of the symmetry of the hyperbolic structure and its directional difference characteristics, designing the corresponding convolution kernel template, and realize effective segmentation of hyperbolic targets in B-SCAN images through convolution operation．
25097 对双曲线目标采用最小二乘法进行二次曲线拟合得到精确的双曲线图像。 The least square method is used to perform quadratic curve fitting on the hyperbolic target to obtain an accurate hyperbolic image．
25098 与基于迁移学习的方法、HOG 算法以及基于 Hough 变换的 B-SCAN 检测算法等相比，本文方法得到的结果在综合指标度量 F 上是最优的。 Compared with B-SCAN im-age detection algorithms such as transfer learning-based methods, HOG (histogram of oriented gradients) algorithm and Hough transform algorithm, the results obtained by the method in this paper are optimal on the comprehensive measurement index F．
25099 针对三维车载自组织网络中，高速移动的车辆节点和复杂多变的链路状态导致车辆间通信链路不稳定的问题，通过引入软件定义网络技术实时获取网络状态并预测其变化过程，构建时-空演化图模型，并定义链路效用指标量化无线链路性能，然后建立基于链路效用的加权时-空演化图模型，最后将路由问题转化为多属性决策问题，设计基于链路效用的可靠路由算法。 In the three-dimensional vehicular adhoc networks (3D-VANET), high-speed moving vehicle nodes and changeable link states lead to unstable inter-vehicle communication links. Aiming at this problem, the time-space evolution graph model is constructed by introducing software define network technology to obtain network state in real-time and predict the process of time change and the link utility index is defined to quantify the wireless link performance. Then the weighted time-space evolution graph model based on link utility is established. Finally, the routing decision-making problem is transformed into a multi-attribute decision-making problem, and a link utility based reliable routing (LUＲＲ)algorithm isdesigned．
25100 仿真结果表明，相对现有四种路由算法，本文所提路由算法在数据包传输率、端到端时延和路由负载率方面，性能均有明显提升。 Simulation results show that, compared with the existing four routing protocols, LUＲＲ algorithm has significantly improved packet transmission rate, end-to-end delay and routing load rate．
25101 在超密集异构无线网络中，针对传统垂直切换算法无法同时描述网络状态的模糊性和随机性，导致网络性能得不到有效提升的问题，提出一种基于区间二型模糊神经网络的垂直切换算法。 In the ultra-dense heterogeneous wireless network, the traditional vertical handoff algorithm can not describe the fuzziness and randomness of the network state at the same time, so the network performance can not be effectively improved. A vertical handoff algorithm based on the interval type II fuzzy neural network is proposed to solve above problem．
25102 重构了两阶段判决算法:在网络预筛选阶段，定义了历史接入率，结合当前候选网络集的数目设置阈值。 A two-stage decision system is reconstructed: in the network's prescreening stage, the historical access rate is defined to set the threshold combine with the number of current candidate network sets．
25103 根据接收信号强度和剩余可用带宽，对用户接收范围内的所有网络进行初步筛选;再在垂直切换判决阶段，将剩余候选网络的时延，丢包率以及误码率作为区间二型模糊神经网络的输入，利用前馈神经网络的结构完成模糊逻辑推理，经训练之后计算得到输出判决值，从而选择最佳接入网络。 According to the received signal strength and the remaining available bandwidth, all the networks within the user's receiving range are preliminarily screened; The delay, packet loss rate and bit error rate of the remaining candidate networks are taken as the inputs of the it2fnn in the vertical handoff decision stage. The fuzzy logic reasoning is completed by using the structure of the feed forward neural network, and the output decision value is calculated after the training, and the optimal network is selected．
25104 实验结果表明，该算法能在保证时间开销较低的同时，有效降低切换决策的错误概率，减少切换失败和切换次数，提升网络总吞吐量。 The simulation results show that the algorithm can ensure low time consumption, and effectively reduce the error probability of handoff decision and the number of handoff failures and handoff times. Meanwhile, it can improve the total throughput of networks．