A Mediator for Approximate Consistency: Supporting “Good Enough” Materialized Views

This paper addresses the needs of application designers whowould like to tell an automated assistant the following: Here is aquery that defines a view I want to materialize within myapplication. I need this view to remain approximately consistent with the state of the data sources from which the view isderived, in accordance with declaratively specified stalenesspredicates. When the view becomes stale, follow the refresh strategyI specify (e.g., eager, lazy, hybrid). You must do this inheterogeneous environments containing both active and passive datasources. This paper describes an architecture that realizes this vision. Theapproach supports materialized, object-based views, called quasi-views,defined over shared databases. Quasi-views arerefreshed according to the consistency conditions and refreshstrategies specified declaratively by application designers. Theseconditions allow for the deviation of quasi-views from their databasecounterparts according to well-defined and monitored approximateconsistency predicates. A layer of software called a Mediator for Approximate Consistencyautomatically generates the databaseobjects necessary to enforce these consistency conditions, shieldingthe application developer from the implementation details ofconsistency maintenance. In addition, it does this for both activeand passive (e.g., legacy) data sources.This paper formalizes quasi-views, presents a declarative quasi-viewspecification language, and describes an architecture andimplementation of a Mediator for Approximate Consistency.     

CCN中一种改进的面向隐私保护的缓存策略

内容中心网络（CCN）是一种新型的网络架构,但其路由器缓存数据包的模式给用户带来了隐私泄露的风险。提出一种面向隐私保护的动态区域协作缓存策略。该策略以兼顾用户隐私保护的同时提高网络性能为目的进行设计,从信息熵的角度出发,以提高用户的请求信息的不确定度为目标,通过将内容存储在相应的隐匿系数高的缓存节点,增加攻击者确定请求用户的难度;以动态区域协作的方式存储内容,增大缓存内容的归属不确定性,以加大攻击者定位数据包的难度。仿真结果表明,该策略可降低内容请求时延,提高缓存命中率。     

Cooperative caching game based on social trust for D2D communication networks

Content sharing via device‐to‐device (D2D) communications has become a promising method to increase system throughput and reduce traffic load. Due to the characteristic of spectrum sharing in D2D network, confidentiality is becoming a key issue in content transmission. Secure communication in D2D networks is generally guaranteed by a physical‐layer security mechanism. However, this method sacrifices the system transmission rate while ensuring security. Since mobile devices are carried by humans, we can leverage their trust relations to enhance the security of communications. As much, considering the psychology structure and social attributes of mobile users, we build a multidimensional trust evaluation mechanism to evaluate the trust relationship between users, and we pick out the trusted users based on the decision‐theoretic rough sets. By sharing content only between trust users, we can enhance the security of content transmissions without relying on physical‐layer security measures. Meanwhile, content caching is now widely used to improve accessing efficiency and reduce traffic load on cellular networks. However, caching content for other users incurs additional cost, which results in selfish and noncooperative behavior in users. Considering such selfishness, we introduce a cooperative caching game based on multidimensional trust relations to motivate users to cache contents for other devices. In this game, the trust relations and physical distance between two users are considered to formulate the cost function. Furthermore, we introduce an incentive caching algorithm based on social trust to minimize the total cost in the D2D network.     

Dynamic theoretic proactive caching placement game for energy saving in H‐CRAN

Edge caching is an effective feature of the next 5G network to guarantee the availability of the service content and a reduced time response for the user. However, the placement of the cache content remains an issue to fully take advantage of edge caching. In this paper, we address the proactive caching problem in Heterogeneous Cloud Radio Access Network (H‐CRAN) from a game theoretic point of view. The problem is formulated as a bargaining game where the remote radio heads (RRHs) dynamically negotiate and decide which content to cache in which RRH under energy saving and cache capacity constraints. The Pareto optimal equilibrium is proved for the cooperative game by the iterative Nash bargaining algorithm. We compare between cooperative and noncooperative proactive caching games and demonstrate how the selfishness of different players can affect the overall system performance. We also showed that our cooperative proactive caching game improves the energy consumption of 40% as compared with noncooperative game and of 68% to no‐game strategy. Moreover, the number of satisfied requests at the RRHs with the proposed cooperative proactive caching scheme is significantly increased.     

Packet‐level‐based traffic aggregation to optimize NDN content delivery

In recent years, named data networking (NDN) has been accepted as the most popular future paradigm and attracted much attention, of which the routing model contains interest forwarding and content delivery. However, interest forwarding is far from the bottleneck of routing optimization; instead, the study on content delivery can greatly promote routing performance. Although many proposals on content delivery have been investigated, they have not considered packet‐level caching and deep traffic aggregation, which goes against the performance optimization of content delivery. In this paper, we propose a packet‐level‐based traffic aggregation (PLTA) scheme to optimize NDN content delivery. At first, the packet format is devised, and data plane development kit (DPDK) is used to ensure same size for each packet. Then, the whole delivery scheme with traffic aggregation consideration is presented. The simulation is driven by the real YouTube dataset over Deltacom, NSFNET, and CERNET topologies, and the experimental results demonstrate that the proposed PLTA has better delivery performance than three baselines in terms of cache hit ratio, delivery delay, network load, and energy efficiency.     

基于增量学习的RocksDB键值系统主动缓存机制

由于分层结构的约束,基于日志结构合并（LSM）树的RocksDB键值存储系统面临着读取性能低下的问题。一种有效的解决方法是对热点数据进行主动缓存,但其面临两个挑战：一是如何在数据分布持续动态变化时对热点数据进行预测,二是如何将主动缓存机制与RocksDB存储结构衔接起来。针对这些挑战,基于预测分析技术,构建了由数据采集、系统交互、系统测试等部分组成的面向RocksDB键值系统的主动缓存框架,能够将热点数据缓存在LSM树的较低层级中;并对数据访问模式进行建模,设计并实现了基于增量学习的热点数据预测分析方法,能够有效减少存储介质的I/O访问次数。实验结果表明该机制能有效提升RocksDB在不同动态工作负载下的数据读取性能。     

基于动态内容流行度的NDN缓存决策和替换策略研究

命名数据网络(NDN)中的路由器节点具有缓存能力,这就极大地提高了网络中的数据发送与检索效率。然而,由于路由器的缓存能力是有限的,设计有效的缓存策略仍然是一项紧迫的任务。为了解决这个问题,提出了一种动态内容流行度缓存决策和替换策略(DPDR)。DPDR综合考虑内容流行度和缓存能力,利用一个和式增加、积式减少(AIMD)的算法动态调节流行度阈值,并将超过流行度阈值的内容存入缓存空间;同时提出了一个缓存替换算法,综合考虑了缓存空间中内容的流行度和内容最后被访问时间等因素,将替换值最小的内容移出内容缓存。大量仿真结果显示,与其他算法相比,本文所提的算法能够有效提高缓存命中率,缩短平均命中距离和网络吞吐量。     

信息中心网络中缓存技术研究综述

因特网中日益增长的内容获取需求促使学术界提出了多种以信息为中心的未来网络架构。这类架构将以主机为中心的通信模式转变为以内容为中心。信息中心网络（ICN）最重要的特征之一是利用内置缓存减少用户获取内容的时延、节省网络带宽和缓解网络拥塞。与传统的内容分发网络（CDN）、对等网络（P2P）和Web缓存系统相比,ICN缓存系统呈现出一系列的新特征。分析了缓存新特征对ICN研究带来的挑战;从多方面重点阐述了ICN缓存的优化方法,详细分析对比了不同缓存策略;指出了未来研究方向并总结全文。     

事件驱动的微服务调用链路数据动态采集方法

微服务调用链路数据是微服务应用系统日常运行中产生的一类重要数据,它以链路形式记录了微服务应用中一次用户请求对应的一系列服务调用信息。由于系统的分布性,微服务调用链路数据产生在不同的微服务部署节点,当前对这些分布数据的采集一般采用全量采集和采样采集两种方法。全量采集会产生较大数据传输和数据存储等成本,而采样采集则可能会漏掉关键的链路数据。因此,提出一种基于事件驱动和流水线采样的微服务调用链路数据动态采集方法,并基于开源软件Zipkin设计实现了一个微服务调用链路数据动态采集系统。该系统首先对不同节点符合预定义事件特征的链路数据进行流水线采样,即数据采集服务端只在某节点产生事件定义的数据时对所有节点采集同一链路数据;同时,针对不同节点的数据产生速率不一致问题,采用基于时间窗口的多线程流式数据处理和数据同步技术实现不同节点的数据采集和传递;最后,针对各节点链路数据到达服务端先后顺序不一的问题,通过时序对齐方式进行全链路数据的同步和汇总。在公开的微服务调用链路数据集上的实验结果表明,相较于全量采集和采样采集方法,所提方法对于包含异常、慢响应等特定事件的链路数据具有采集准确性高、效率好的效果。     

D2D协作边缘缓存系统中基于传输时延的缓存策略

为了满足5G系统低时延高可靠的需求,针对单缓存终端直传(D2D)协作边缘缓存系统,提出了一种基于传输时延的缓存策略。运用随机几何理论,将请求用户和空闲用户的动态分布建模为相互独立的齐次泊松点过程,综合考虑内容流行度、用户位置信息、设备传输功率以及干扰,推导出用户的平均传输时延与缓存概率分布的关系式。以平均传输时延为目标函数建立优化问题,提出了一个低复杂度的迭代算法,得到平均传输时延次优的缓存策略。仿真结果表明,该缓存策略在传输时延方面优于常见的几种缓存策略。