论文标题
用于边缘缓存的全双工收音机
Full-Duplex Radios for Edge Caching
论文作者
论文摘要
本章重点是在超密集网络(UDNS)的背景下将缓存功能添加到全双制(FD)无线电上带来的性能增强。更具体地说,我们的目的是表明此类网络的干扰足迹,即,由于Edge Caching,可以大大降低该网络的主要瓶颈,以观察网络级别的理论FD吞吐量加倍。基本结果表明,与其半双链(HD)对应物相比,此类网络的大部分收益只有在执行其基础架构的昂贵修改和/或当用户设备之间(UES)之间在适当的控制链接上交换高速率信号时,才能实现此类网络。因此,我们旨在根据网络边缘的当地预先提取的本地内容提出可行且具有成本效益的替代方案。我们首先考虑了一种干扰的场景,例如超密集的FD小细胞网络,其中几个非合作的FD基站(BSS)在与无线回程节点(BN)通信以检索内容以交付的过程中为其相关的UES提供服务。然后,我们描述旨在捕获本地文件流行的地理缓存策略,并计算相应的缓存概率。因此,我们计算了UE要求成功传输文件的概率,即直接通过其服务小细胞基站(SB)或相应的BN传输文件:然后使用该数量来降低所考虑网络的吞吐量。我们的方法利用随机几何形状的工具,以确保问题的分析性和结果的普遍性。我们的数值模拟表明,从无缓存到缓存的FD小型细胞网络的转移可以改善性能。
This chapter focuses on the performance enhancement brought by the addition of caching capabilities to full-duplex (FD) radios in the context of ultra-dense networks (UDNs). More specifically, we aim at showing that the interference footprint of such networks, i.e., the major bottleneck to overcome to observe the theoretical FD throughput doubling at the network level, can be significantly reduced thanks to edge caching. Fundamental results show that most of the gain, as compared to their half-duplex (HD) counterparts, can be achieved by such networks only if costly modifications to their infrastructure are performed and/or if high-rate signaling is exchanged between user equipments (UEs) over suitable control links. Therefore, we aim at proposing a viable and cost-effective alternative to these solutions based on pre-fetching locally popular contents at the network edge. We start by considering an interference-rich scenario such as an ultra-dense FD small-cell network, in which several non-cooperative FD base stations (BSs) serve their associated UEs while communicating with a wireless backhaul node (BN) to retrieve the content to deliver. We then describe a geographical caching policy aiming at capturing local files popularity and compute the corresponding cache-hit probability. Thereupon, we calculate the probability of successful transmission of a file requested by a UE, either directly by its serving small-cell base station (SBS) or by the corresponding BN: this quantity is then used to lower-bound the throughput of the considered network. Our approach leverages tools from stochastic geometry in order to guarantee both analytical tractability of the problem and generality of the results. Our numerical simulations show that shifting from cache-free to cache-aided FD small-cell networks yields a remarkable performance improvement.