5G ağlarında sosyal grup bilgilerine dayalı veri önbellekleme yöntemi

Increased demand for media content by mobile applications has imposed huge pressure on wireless cellular networks to deliver the content efficiently and effectively. To keep up with this demand, mobile edge computing (MEC), also called multi-access edge computing, is introduced to bring cloud computing and storage capabilities to the edges of the cellular networks, such as 5G, with the aim of increasing quality of service to applications and reducing network traffic load. One important application of multi-access edge computing is data caching. As significant portion of multimedia data traffic is generated from media sharing and social network services, various mobile edge caching schemes have emerged to improve the latency performance of these applications. In this thesis, driven from the fact that social interaction between mobile users has a strong influence on data delivery patterns in the network, we propose a socially-aware edge caching system model and methods that consider social groups of users in caching decisions together with storage and transmission capacities of edge servers. Unlike other studies, where users are manually grouped according to their interests, our approach is based on user-specified social groups, where users in a group are neither obligated to share the same interests nor be attentive to the shared content. Our methods cache content considering locations of members of social groups and the willingness of these members in using the related applications. We evaluate the performance of our proposed methods with extensive simulation experiments. The results show that our methods can significantly reduce user-experienced latency and network load.