陈炳才

个人信息Personal Information

教授

博士生导师

硕士生导师

性别:男

毕业院校:哈尔滨工业大学

学位:博士

所在单位:计算机科学与技术学院

学科:计算机应用技术. 通信与信息系统

办公地点:创新园大厦

联系方式:手机:15504280859; 微信:33682049;

电子邮箱:china@dlut.edu.cn

扫描关注

论文成果

当前位置: 中文主页 >> 科学研究 >> 论文成果

Caching D2D Connections in Small-Cell Networks

点击次数:

论文类型:期刊论文

发表时间:2018-12-01

发表刊物:IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY

收录刊物:SCIE、Scopus

卷号:67

期号:12

页面范围:12326-12338

ISSN号:0018-9545

关键字:Bipartite graph theory; caching; device-to-device; Kuhn-Munkres algorithm; link scheduling; small-cell networks

摘要:Small-cell network is a promising solution to high video traffic. However, with the increasing number of devices, it cannot meet the requirements from all users. Thus, we propose a caching device-to-device (D2D) scheme for small-cell networks, in which caching placement and D2D establishment are combined. In this scheme, a limited cache is equipped at each user, and the popular files can be prefetched at the local cache during off-peak period. Thus, dense D2D connections can be established during peak time aided by these cached users, which will reduce the backhaul pressure significantly. To do this, first, an optimal caching scheme is formulated according to the popularity to maximize the total offloading probability of the D2D system. Thus, most edge users can obtain their required video files from the caches at users nearby, instead from the small-cell base station. Then, the sum rate of D2D links is analyzed in different signal-to-noise ratio (SNR) regions. Furthermore, to maximize the throughput of D2D links with low complexity, three D2D-link scheduling schemes are proposed with the help of bipartite graph theory and Kuhn-Munkres algorithm for low, high, and medium SNRs, respectively. Simulation results are presented to show the effectiveness of the proposed scheme.