应用于协议无感知转发交换机的流缓存方法

协议无感知转发支持任意协议的解析和处理,增强了软件定义网络的可编程能力。为提高转发性能,该文提出一种应用于协议无感知转发交换机的流缓存方法,通过识别匹配和动作的依赖关系,得到匹配字段的绝对位置,用以预先解析报文。为确保流缓存的加速效果,根据匹配类型与表项数量选择应用流缓存的流表。此外,该文对比了单流表缓存与多流表缓存对转发性能的提升,并提出了根据网络流量实际情况的自适应切换策略。通过扩展POFSwitch实现所提方法,并用实际规则与骨干网流量进行验证,应用流缓存后,交换机报文转发速率提升了220%。流缓存可以为可编程数据平面提供更高的转发性能。     

车联网中基于信息年龄价值的边缘缓存策略

随着第五代移动通信（5G）技术的快速发展,车联网（Internet of Vehicles,IoV）已成为5G的重要应用场景之一.针对IoV应用对即时性内容缓存的迫切需求,本文研究了支持动态内容缓存的移动边缘缓存方案以确保动态内容的时效性,该方案允许车辆从边缘缓存接入点（Caching-enabled Access Point,CAP）获得具有新鲜度约束的缓存内容.此外,本文基于协作缓存,利用概率论建立了CAP缓存刷新和交付的随机模型.同时,以基于信息年龄（Age of Information,AoI）的缓存命中率为IoV缓存的服务质量（Quality of Service,QoS）指标,建立了系统缓存命中的价值模型.为确保缓存内容的有效性,本文在分层切片框架中考虑了两种具有不同AoI要求的车辆内容请求.最后,本文提出了一个基于协作内容放置的系统内容缓存命中价值最大化的优化问题,这是一个NP-hard多目标多维多选择背包问题.因此,本文将其划分为若干个相互独立的优化问题并设计了一种基于内容年龄价值的最大增量算法进行求解.仿真结果表明,本文采用的CAP协作内容放置策略的性能优于其他传统方...     

信息中心网络中基于节点缓存命中贡献率的缓存替换方案

信息中心网络(CCN)是一种全新的网络架构,其显著的特点是处处缓存,合理的内容缓存部署能够显著提高网络传输效率.缓存替换策略是缓存管理中的重要组成部分,合理地进行缓存内容的替换,成为影响网络整体性能的关键.考虑到内容自身的特性,设计了一种基于节点缓存命中贡献率的贪婪双倍命中(GDH)缓存替换方案.该方案综合考虑了内容的请求次数、传输代价、缓存代价,设计全新的多目标价值函数,用于评估内容的缓存价值,当缓存空间不足时,替换掉价值最小的内容,实现节点缓存内容价值的最大化.仿真结果表明,该替换算法提高了节点的命中率,降低了获取内容的平均跳数.     

基于网络代价的流媒体缓存策略研究

该文构造了一种新的流媒体缓存效用函数,该函数综合考虑流媒体节目的流行度特性及传输网络的代价参数;设计了一种针对多视频服务器、基于网络代价的流媒体缓存分配与替换算法(Network Cost Based cache allocation and replacement algorithm, NCB)。仿真实验结果显示,NCB算法有效提高了缓存命中率,降低了传送流媒体所消耗的总体网络代价;该算法在网络结构复杂、节目数量庞大的Internet流媒体应用环境中表现出较优越的性能。     

基于社会网络洪泛算法的消息队列转发策略设计

随着智慧城市的发展,移动智能终端的普及,社会网络成为研究人员热议的话题。由于洪泛算法在网络投递率、网络延迟方面有着较为突出的表现,文章主要基于社会网络的洪泛算法,设计了MUOB的消息队列转发策略。通过实验结果表明MUOB转发策略在网络投递率、网络延迟方面优于传统的随机转发策略,先进先出转发策略以及最近的Min Hop转发策略。     

车联网场景下基于内容中心网络的优化缓存策略

车联网中,如何有效选择缓存位置和缓存内容对于提高整体网络性能至关重要。针对上述问题,引入了内容中心网络技术,提出了一种新的优化缓存策略——缓存位置和缓存内容的选择取决于车辆节点值和内容流行度(Vehicle Node Value and Content Popularity,VNVCP)。首先,定义了连通性、中间中心性和特征向量中心性3个车辆节点属性用来评估车辆节点的值,具有不同值的车辆节点缓存具有不同流行度的内容,内容的重要性由其受欢迎程度决定。其次,该策略利用不同类型内容受欢迎程度的差异确保缓存内容分布均匀,同时评估来自多个属性的车辆节点的值以提高车辆节点利用率。仿真结果表明,VNVCP在缓存命中率、平均跳数和传输延迟方面明显优于传统的LCE(Leave Copy Every where)、Prob(0.5)和MPC(Most Popular Content)。     

面向分域信息中心网络的协作混合缓存和路由策略

网内缓存技术一直是信息中心网络(Information-Centric Networking, ICN)研究中的核心问题。传统缓存方案通常为网络内的单一全局协作策略,然而不同网络区域的缓存目标不同,导致传统方案的性能优化程度有限,可扩展性不强。同时协同缓存放置和请求路由,相较于纯缓存方案有效性更高。该文提出并评估了一种新型面向分域的协作混合缓存放置和路由请求机制,该机制将缓存网络区分核心网络和边缘网络,在核心网络中采用非沿路的HASH协作机制,边缘网络采用沿路的回退协作机制,创建二元组标签保存缓存放置的信息并作用于后续的请求转发。通过仿真对比表明,该文提出的协作混合缓存机制相较于传统策略可以更好地平衡网络资源利用和用户体验,并在大型ISPs中拓展性更好。     

信息中心网络中网络缓存的角色探索

作为信息中心网络(information-centric networking,ICN)的特色之一,网络缓存在彰显ICN的优势方面扮演着重要的角色,但当前对网络缓存所扮演的具体角色没有一致的看法。指出网络缓存应具有三级缓存,解释这三级网络缓存可以在错误恢复、突发访问缓解、DDoS攻击防治、热点内容缓存及托管服务方面起一定的作用,并说明了实现这些角色需要解决的问题。     

基于压缩感知的车联网数据采集

当前车联网已经成为了大规模城市交通信息监测的重要平台。如何将车辆中采集的海量数据同时传输给监测中心并进行处理却是急需解决的问题。通过研究中国深圳市约3000辆出租车采集的数据发现,这些数据存在着空间相关性。基于以上的发现,提出了一种新的应用在车联网交通监测的压缩感知算法,在保证恢复精确度的同时大量减少采样所需要的数据量。仿真实验表明在较低的采样率情况下,该算法重构信号的误差相对很低。     

面向5G-ICN切片的分层缓存策略

多接入边缘计算(Multiple Access Edge Computing,MEC)和网络切片在5G网络中扮演了重要角色。网络切片的灵活可编程性和MEC的去中心化促进了信息中心网络(Information Centric Network,ICN)的部署。但是,ICN中缓存资源分配的位置和数量严重影响了运营商的收益。为此,首先将移动虚拟网络运营商(Mobile Virtual Network Operator,MVNO)和多个内容提供商(Content Provider,CP)建模为一主多从的Stackelberg博弈;其次,将MVNO内部的缓存资源分配建模为一个联合缓存成本和传输能耗的混合整数规划问题,并利用改进的精英遗传算法得到最佳位置;最后,通过博弈迭代算法得到运营商的最优定价和CP的最优放置数量。仿真结果表明,该算法具有较高的有效性以及较好的缓存资源分配能力,能够降低传输能耗和缓存成本,与贪婪搜索(Iteration Greedy Search,IGS)算法、联合缓存分配和延时控制(Joint Cache Allocation and Delay Control,JCADC)算法相比传输能耗分别节省了5%、11%,运营商的收益分别提高了32%、44%。     

基于收益感知的信息中心网络缓存机制

针对信息中心网络（ICN）内置缓存系统中的海量内容块流行度获取和存储资源高效利用问题,以最大化节省内容访问总代价为目标,建立针对内容块流行度的缓存收益优化模型,提出了一种基于收益感知的缓存机制。该机制利用缓存对请求流的过滤效应,在最大化单点缓存收益的同时潜在地实现节点间协作和多样化缓存;使用基于布隆过滤器的滑动窗口策略,在检测请求到达间隔时间的同时兼顾从源服务器获取内容的代价,捕获缓存收益高的内容块。分析表明,该方法能够大幅压缩获取内容流行度的存储空间开销;仿真结果表明,该方法能够较为准确地实现基于流行度的缓存收益感知,且在内容流行度动态变化的情况下,在带宽节省和缓存命中率方面更具优势。     

A QoE centric distributed caching approach for vehicular video streaming in cellular networks

Distributed caching‐empowered wireless networks can greatly improve the efficiency of data storage and transmission and thereby the users' quality of experience (QoE). However, how this technology can alleviate the network access pressure while ensuring the consistency of content delivery is still an open question, especially in the case where the users are in fast motion. Therefore, in this paper, we investigate the caching issue emerging from a forthcoming scenario where vehicular video streaming is performed under cellular networks. Specifically, a QoE centric distributed caching approach is proposed to fulfill as many users' requests as possible, considering the limited caching space of base stations and basic user experience guarantee. Firstly, a QoE evaluation model is established using verified empirical data. Also, the mathematic relationship between the streaming bit rate and actual storage space is developed. Then, the distributed caching management for vehicular video streaming is formulated as a constrained optimization problem and solved with the generalized–reduced gradient method. Simulation results indicate that our approach can improve the users' satisfaction ratio by up to 40%. Copyright © 2015 John Wiley & Sons, Ltd.     

基于概率的启发式ICN缓存内容放置方法

提出一种基于概率存储的启发式住处中心网络内容缓存方法（PCP）。主要思想是请求消息和数据消息在传输过程中统计必要信息,当数据消息返回时,沿途各缓存节点按照一定概率决策是否在本地缓存该内容。设计缓存概率时综合考虑内容热度和缓存放置收益,即内容热度越高,放置收益越大的内容被缓存的概率越高。实验结果表明,PCP在缓存服务率、缓存命中率、平均访问延迟率等方面,与现有方法相比具有显著优势,同时PCP开销较小。     

基于可重构路由器上缓存的流媒体协作分发策略

提出一种新颖的基于可重构路由器上缓存的协作分发策略来加速流媒体。通过网络存储即多个边缘路由器节点对热点视频数据进行合作缓存,就近为用户提供服务,从而使得流媒体服务器的性能要求尤其是带宽需求得到巨大的降低,骨干网传输的流量也明显减少,同时用户响应延迟也得到明显的改善。此外,实现了一个原型系统来评价基于路由器上缓存的流媒体协作分发策略的性能,结果表明该方案相比于现有的方案在改善网络性能以及用户体验方面取得很大的提升。     

流媒体代理缓存技术的研究

随着Internet技术的发展和普遍应用,流媒体技术在Internet上得到了广泛的应用.对流媒体对象的访问,需要高且稳定的传送速率,网络带宽消耗大且持续时间长,容易给其他类型文件的访问带来影响,若用户过多,还会使初始流媒体服务器过载.代理缓存技术可帮助解决上述问题.文中介绍了流媒体代理缓存的特点,流媒体代理缓存的算法,流媒体代理缓存的评价指标和影响流媒体代理缓存效果的因素.     

Tile caching for scalable VR video streaming over 5G mobile networks

Currently, VR video delivery over 5G systems is still a very complicated endeavor. One of the major challenges for VR video streaming is the expectations for low latency that current mobile networks can hardly meet. Network caching can reduce the content delivery latency efficiently. However, current caching schemes cannot obtain ideal results for VR video since it requests the viewport interactively. In this paper, we propose a tiled scalable VR video caching scheme over 5G networks. VR chunks are first encoded into multi-granularity quality layers, and are then partitioned into tiles to facilitate viewport data access. By accommodating the 5G network infrastructure, the tiles are cooperatively cached in a three-level hierarchal system to reduce delivery latency. Furthermore, a quality-adaptive request routing algorithm is designed to cater for the 5G bandwidth dynamics. Experimental results show that the proposed scheme can reduce the transmission latency over conventional constant bitrate video caching schemes.     

面向ICN的动态自适应流媒体算法

信息中心网络(information-centric networking,ICN)与传统IP网络的重要区别在于,前者的路由器节点具备缓存能力,能将热门内容推送到更靠近客户的网络边缘,从而降低客户获取内容的时延。然而ICN中的节点缓存特性会导致传统的自适应流媒体协议出现视频抖动甚至播放中断,降低客户的观看体验。提出了ICN中的动态自适应流媒体算法,在保证客户端播放流畅的同时,尽可能地提升视频片段的码率,提升客户的观看体验。通过在现有的ICN平台中进行仿真实验,证明了该自适应算法在性能上优于传统的算法。     

Link efficiency and quality of experience aware routing protocol to improve video streaming in urban VANETs

In a vehicular ad‐hoc network (VANET), vehicles can play an essential role in monitoring areas of a smart city by transmitting data or multimedia content of environmental circumstances like disasters or road conditions. Multimedia content communication with quality of experience (QoE) guarantees is a challenging undertaking in an environment such as that of a VANET. Indeed, a VANET is characterized by numerous varying conditions, significantly impacting its topology, quality of communication channels, and paths with respect to bandwidth, loss, and delay. This paper introduces a link efficiency and quality of experience aware routing protocol (LEQRV) to improve video streaming provisioning in urban vehicular ad‐hoc networks. LEQRV uses an enhanced greedy forwarding‐based approach to create and maintain stable high quality routes for video streaming delivery. It improves the performance of the quality of experience by increasing the achieved QoE scores and reducing the forwarding end‐to‐end delay and frame loss.     

Incentive analysis for cooperative interactive multiview video streaming

In interactive multiview video streaming (IMVS), users can periodically select one out of many captured views available for observation. In single-view video streaming, cooperative strategies where peers share received packets of the same video have proven to be effective in reducing server׳s upload burden, and incentive mechanisms are designed to stimulate user cooperation. However, exploiting user cooperation in higher dimensional IMVS is difficult, since users watching different views makes it difficult to establish partnership, and users switching views frequently and independently makes it difficult to maintain partnership over time. In this paper, we use a multiview video frame structure for IMVS to support cooperative view-switching, where peers may help each other even if they are observing different views. We then model peers׳ interaction as an indirect reciprocity game, where each user is assigned a reputation level. To gain a higher reputation level, users help others, which in turn leads to a higher likelihood to receive others׳ help later. In this work, we focus on how view switching, the key feature of IMVS, affects user cooperation. By modeling users׳ decision making as a Markov decision process, our analysis shows that users tend to cooperate at some views but not others: given peers can predict their future view navigation paths probabilistically, for a peer who is likely to enter a view-switching path not requiring others׳ help, he also has less incentive to cooperate. Furthermore, we observe that the game may have multiple Nash Equilibria corresponding to different cooperation levels, e.g., users cooperate at all views in the full cooperation equilibrium, while users only cooperate at certain views in the partial cooperation equilibrium. The particular equilibrium the game will converge to depends on the initial cooperation level of the game. To stimulate user cooperation at all views, we propose a Pay-for-Cooperation (PfC) scheme at the beginning of the game to drive the game to the full cooperation equilibrium to improve system efficiency. Our simulation results show the effectiveness of PfC.