Obviously, introducing cache function in routers brings many benefits. However, how to design its strategies to further improve caching efficiency is a major challenge for Information-Centric Networking (ICN), where many problems still remain to be solved. Firstly, the complexity of the existing schemes is usually high, which makes it difficult to meet the requirements of wire-speed forwarding of routers. In addition, they are not designed for time-varying/mutable content data, resulting in the mass storage of outdated content data in the network and waste of limited in-network caching resource. Secondly, in terms of verification methods, the existing simulators do not consider the impact of cache internal memories on the access performance of content data, and cannot simulate their characteristics in processing delay and throughput, which also leads to deviations in performance evaluation of caching strategies. Therefore, in this project, we firstly propose a router caching cooperation strategy with low complexity and overhead, to achieve effective identification, edge caching, nearby-access and collaboration caching of popular content data in the network. Secondly, we propose an efficient and low-cost strategy to guarantee caching consistency, with effective deletion and update of outdated content data cached in routers. Thirdly, we design a hardware-level cache simulator to simulate caching access characteristics, which contributes to evaluate the cache performance more realistically. The project aims to fully enhance ICN caching efficiency and provide technical support for its popularization and application.
路由器引入缓存功能的好处显而易见,然而,如何设计缓存策略以提高其效益是信息为中心网络(ICN)所面临的重大挑战,尚存诸多问题亟待解决。首先,现有方案操作复杂度通常较高,难以满足路由器线速转发需求;且未设计对时变内容数据的处理机制,造成失效内容数据在网络中大量存储、浪费有限缓存资源。其次,在验证手段方面,现有模拟器并未考虑缓存内部存储器对内容数据存取性能的影响,无法模拟处理时延、吞吐量等特性,也将导致其对缓存策略性能评估产生偏差。为此,本项目首先提出低复杂度、低开销的路由器缓存协作策略,实现网络对流行内容数据的有效识别、边缘缓存、就近获取和协同存储;其次,提出高效的路由器缓存一致性保障策略,在控制信令开销的同时,实现路由器对过期内容数据的及时删除和更新;第三,设计硬件层面缓存模拟器,实现缓存器存取特性的模拟,以更加真实地评估缓存性能。项目旨在全面提升ICN缓存效益,为其推广应用提供技术支撑。
本项目围绕“以信息为中心网络缓存协同优化策略与验证方法”开展以下三方面研究:(1)轻量级缓存协作策略,实现网络对流行内容数据的有效识别、边缘缓存、就近获取以及协同存储,提升网络传输性效率;(2)高效的缓存一致性保障策略,实现路由器对过期内容数据的有效删除与及时更新的同时,降低相关信令开销,提升缓存利用率;(3)缓存模块模拟器设计,实现同构或异构缓存介质组合存取特性的模拟,以更加真实地评估缓存效能。此外,研究基于网络缓存的可靠传输机制;探索缓存在卫星网络、边缘计算等新兴领域的应用场景并对实际效果进行评估。发表(含录用)期刊论文13篇、会议论文3篇;获授权专利7项;已完成项目既定目标。相关成果可为信息为中心网络、网络缓存的实际应用提供理论基础与技术参考。
{{i.achievement_title}}
数据更新时间:2023-05-31
跨社交网络用户对齐技术综述
城市轨道交通车站火灾情况下客流疏散能力评价
基于FTA-BN模型的页岩气井口装置失效概率分析
基于图卷积网络的归纳式微博谣言检测新方法
多源数据驱动CNN-GRU模型的公交客流量分类预测
以信息为中心的网络(ICN)缓存机制性能评估与算法优化
信息中心网络中缓存策略与激励机制研究
信息为中心的天地一体化网络星间路由机制与缓存管理策略研究
面向多自治域环境的信息中心网络缓存策略和机制研究