Content Distribution In 5G Mobile Communication Networks

With the rapid development of mobile communication systems, mobile devices have been extensively emerging up in recent years. Future mobile communications(5G) face enormous challenges as traditional voice services are replaced with increasing mobile multimedia and data services. To address the vast data traffic volume and the requirement of user Quality of Experience(QoE), and reduce the cost of operation in the next generation mobile networks, it is imperative to develop efficient content distribution technique, aiming at significantly reducing redundant data transmissions and improving mobile network performance. On the other hand, in recent years cloud computing as a promising new content-centric paradigm is exploited to fulfill the service requirements by provisioning data storage and computing resources on demand.However, due to the centralized architecture of current cellular networks, the capability and resources of the wireless link, radio access network, mobile backhaul and mobile core network cannot effectively cope with explosively growing mobile traffic. To address the demand of vast data traffic volume on future communication systems, in this MEng thesis we conduct following research on content caching techniques:First, we conduct a comprehensive survey study on the key techniques of content distribution and caching for 5G systems. Considering the general consensus about 5G systems, and on the base of previous research work, we propose Hierarchical Cloud Service Networks(HDCSN) architecture, which is mainly composed of three parts: Cloud Radio Access Networks(C-RAN), Distributed Cloud Service Networks(DCSN), and Core Cloud Data Centers(DCs). Cloud service providers deploy a plurality of Cloudlets in the network forming a DCSN, and pre-allocate content services in local cloudlets from DCs to avoid redundant content transmissions.Next, we propose a cooperative caching framework in HDCSN, where Cloudlets share data content by collaborating with each other. Based on this framework, we propose a cooperative content distribution optimization algorithm, which uses content popularity to characterize content. The content state, which is exploited to dynamically manage caches, is determined by content requests, editorial updates and new arrivals. Based on content state, we formulate a content distribution optimization model to obtain an optimal content distribution solution, aiming at minimize the average total transmission delay in HDCSN. As illustrated by computer simulation experiments, the proposed framework can effectively improve content cache hit rate, reduce content delivery latency and outbound traffic volume.To implement cache updates and minimize transmission cost in networks, we futher propose a cache replacement and data transmission strategy based on Markov Decision Process(MDP). The proposed strategy in HDCSN uses random linear network coding to encode the service data(e.g. web page, videos) in DCs and uses popularity to characterize various data. We propose three metric standards for Knapsack Problem in Greedy Algorithm, which are exploited to make dynamic decisions on cache replacement and data transmission based on MDP, aiming at minimizing the average content transmission cost in HDCSN. Simulation results show that the proposed strategy can significantly improve the cache hit rate as well as reduce average transmission cost in HDCSN.