基于路径代价和节点代价的信息中心网络缓存策略

发布时间：2018-03-27 13:39

本文选题：信息中心网络　切入点：缓存网络　出处：《小型微型计算机系统》2017年11期

【摘要】：网络化缓存是ICN(信息中心网络)架构的重要特征之一,对改善网络性能起着重要作用,如何合理利用有限的缓存空间,在优化缓存部署时实现缓存开销最小是一个值得关注的问题.考虑到缓存开销不仅包含节点访问缓存节点时的路径代价,还应包含被新缓存内容所替换的旧内容的替换代价,提出了基于路径访问代价和节点替换代价的缓存策略(Path and Node Cost based Caching Strategy,简称PNCCS).该文首先建立了一个理论模型来分析缓存部署对路径访问代价和节点替换代价的影响.基于这个模型,缓存部署问题可以形式化地描述成一个最优化问题,并采用离散粒子群优化算法进行求解,最优解代表着一种优化的缓存部署方案.实验结果表明,PNCCS策略与CEE、Prob和LCD相比,在降低网络运行开销的同时提升了用户体验质量两方面的性能指标.
[Abstract]:Network caching is one of the important features of ICN (Information Center Network) architecture, which plays an important role in improving network performance. Implementing minimum cache overhead in optimizing cache deployment is a matter of concern. Considering that cache overhead includes not only the path cost for nodes to access cached nodes, but also the replacement cost for old content replaced by new cached content. This paper proposes a cache policy based on path access cost and node replacement cost: path and Node Cost based Caching strategy. In this paper, we first establish a theoretical model to analyze the influence of cache deployment on path access cost and node replacement cost. Based on this model, The cache deployment problem can be formally described as an optimization problem and solved by discrete particle swarm optimization (DPSO). The optimal solution represents an optimized cache deployment scheme. Experimental results show that the PNCCS strategy is compared with CEEN Prob and LCD. At the same time, it improves the performance index of user experience quality.
【作者单位】：东北大学秦皇岛分校控制工程学院;东北大学信息科学与工程学院;东北大学软件学院;
【基金】：河北省高等学校科学技术研究项目(QN2014327)资助国家杰出青年科学基金项目(61225012,71325002)资助
【分类号】：TP393.02

【相似文献】