内容中心网络缓存策略的研究

发布时间：2018-11-28 20:12

【摘要】：传统的基于IP地址的传输模式存在大量的重复内容传输,为此,内容中心网络(Content-Centric Network,CCN)提出采用数据命名路由协议,通过路由器缓存内容减少重复传输,提高传输效率。然而,网络中缓存容量远小于全网内容,缓存每一个经过的内容会造成极大的缓存冗余和频繁的缓存替换。因此,如何高效利用缓存资源是内容中心网络亟需解决的关键问题。本文从缓存系统架构、用户请求和无线接入等角度出发,着重研究内容中心网络缓存决策问题。本文主要研究工作如下:1.研究用户请求分布不均衡情况下的缓存决策问题。本文首先提出基于软件定义内容中心网络(Software Defined Content Centric Network,SDCCN)的缓存决策系统框架,并对其中的缓存决策问题进行描述和难度分析。然后,以最小化所有用户请求时延为目标建立整数线性规划模型(CDP-ILP)从而提供问题的最优解。为了快速求解,提出了三种快速的启发式算法——AlgRelax,AlgGreedy,AlgMix。最后,通过自主设计的内容中心网络网内缓存系统仿真实验平台,实现内容中心网络的转发机制以及文中提出的算法和已有的缓存策略。从用户请求总时延,缓存命中次数,服务器负载三个方面进行算法对比,仿真结果表明提出的缓存决策算法能够减少用户请求总时延,增加缓存命中次数,降低服务器负载。2.进一步考虑了无线用户可以选择接入路由器的情况。同时优化路由器选择和缓存决策,以最小化用户请求的总时延。为此建立二次规划模型,并提出一种基于请求相似度的用户接入算法——AlgUACC。随后对缓存决策仿真实验平台进行扩展,加入无线用户接入模块,实现基于请求相似度的用户接入算法以及随机接入算法。实验结果表明,提出的算法相比随机接入算法能够减少用户的请求时延。
[Abstract]:There is a lot of duplicate content transmission in the traditional IP address transmission mode. For this reason, the content Center Network (Content-Centric Network,CCN) proposes a data named routing protocol to reduce the repeat transmission through the content cache of the router. Improve transmission efficiency. However, the cache capacity in the network is much smaller than that in the whole network. Therefore, how to make efficient use of cache resources is a key problem that needs to be solved in content-centric networks. Based on the architecture of cache system, user request and wireless access, this paper focuses on the content-centric network cache decision-making. The main work of this paper is as follows: 1. In this paper, the cache decision problem in the case of uneven distribution of user requests is studied. In this paper, a framework of cache decision system based on software defined content-centric network (Software Defined Content Centric Network,SDCCN) is proposed, and the cache decision problem is described and analyzed. Then, an integer linear programming model (CDP-ILP) is established to provide the optimal solution to the problem, aiming at minimizing the delay of all user requests. In order to solve the problem quickly, three fast heuristic algorithms, AlgRelax,AlgGreedy,AlgMix., are proposed. Finally, the forwarding mechanism, algorithms and cache strategies of the content-centric network are realized by using the self-designed simulation platform of the cache system in the content-centric network. Compared with the three aspects of user request total delay, cache hit times and server load, the simulation results show that the proposed cache decision algorithm can reduce the total delay of user requests and increase the number of cache hits. Lower server load. 2. Further consideration is given to the situation in which wireless users can choose access routers. At the same time, the router selection and cache decision are optimized to minimize the total delay of user requests. In this paper, a quadratic programming model is established, and a user access algorithm based on request similarity, AlgUACC., is proposed. Then the buffer decision simulation experiment platform is extended to add the wireless user access module to realize the user access algorithm based on request similarity and the random access algorithm. Experimental results show that the proposed algorithm can reduce the request delay compared with the random access algorithm.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.0

【参考文献】