异构网络中的编码缓存性能分析

As the network traffic continues to grow intensively, caching plays a more and more important role in reducing the network traffic. Recently, a new class of caching schemes, called coded caching, has received a significant amount of interest. This is because coded caching can exploit new multicast capabilities even when multiple users request different pieces of content. While most of the prior studies of coded caching have focused on homogeneous settings, the performance limits of coded caching under heterogeneous system settings is less investigated. Therefore, this project aims to study coded caching under heterogeneous settings, i.e., file popularity, different application requirements, and various transmission topologies. The main contributions are summarized as follows..1).To design the caching strategy based on the theoretical threshold of content popularity, demonstrate the impact of contents with low popularity, and to achieve constant gap between achievable rate and the theoretical lower bound..2).To design the caching strategy in which a content will be cached quadratically proportional to its rate requirement, construct a new set of request patterns, and using entropy inequalities to decrease the transmission rate needed..3).To propose cooperate transmission strategies between users in different layers, investigate the impact of channel fading in wireless networks and the complex topologies in wireline networks, design low complexity schemes to increase the caching efficiency..Overall, this project is to provide useful and insightful guidelines for the use of coded caching in practical networks.

随着网络流量持续快速的增加，存储或者缓存在降低网络负载方面的重要性日益突出。近年来，一类新的缓存算法，叫做编码缓存受到越来越多的关注。这是因为编码缓存能够有效利用多播机会，即使不同用户请求的内容不一样。现有研究很大一部分关注的是同构网络下的性能分析，对于异构网络的编码缓存设计和分析目前还比较欠缺。因此，本项目拟从内容流行度、多业务、传输网异构等角度对编码缓存性能进行分析。具体包括：（1）设计基于内容流行度阈值的缓存策略，解决低流行度文件难以刻画的问题，从而达到在任意网络规模任意内容流行度下传输速率与理论下界只有常数差距。（2）提出适应多业务特征的平方比缓存策略，构造新的用户请求模式集合，利用信息论不等式降低服务所需速率。（3）提出跨层的合作传输策略，解决无线信道衰落和有线网络拓扑复杂问题，设计低复杂度算法，提高缓存利用效率。研究成果将为编码缓存在实际网络中的应用提供有益参考和借鉴。

随着网络流量持续快速的增加，存储或者缓存在降低网络负载方面的重要性日益突出。近年来，一类新的缓存算法，叫做编码缓存受到越来越多的关注。这是因为编码缓存能够有效利用多播机会，即使不同用户请求的内容不一样。现有研究很大一部分关注的是同构网络下的性能分析，对于异构网络的编码缓存设计和分析目前还比较欠缺。因此，本项目拟从内容流行度、多业务、传输网异构等角度对编码缓存性能进行分析。具体包括：（1）设计基于内容流行度阈值的缓存策略，解决低流行度文件难以刻画的问题，从而达到在任意网络规模任意内容流行度下传输速率与理论下界只有常数差距。（2）提出适应多业务特征的平方比缓存策略，构造新的用户请求模式集合，利用信息论不等式降低服务所需速率。（3）提出跨层的合作传输策略，解决无线信道衰落和有线网络拓扑复杂问题，设计低复杂度算法，提高缓存利用效率。研究成果将为编码缓存在实际网络中的应用提供有益参考和借鉴。