P2C2-Popular Content Prediction and Collaboration in mobile edge caching

摘要

Evolution from todays host-centric network architecture to a data-centric network architecture, named data networking (NDN) is a proposed Future Internet architecture based on content name-oriented that pushes and caches data to edge gateways/routers. If the content is popular, the previous queried content can be reused multiple times and it should be kept in the limited size of storage longer than unpopular contents. In the context of mobile data networks, researchers conclude that NDN is one of the best fit solutions in mobile edge caching. Thus, it is critical to study an efficient replacement policy to achieve a higher hitting rate than the original policies applied in NDN protocol. This paper proposes a novel Popular Content Prediction and Collaboration (P2C2) in the context of mobile edge caching. OPNET simulation tool is used to conduct long term evolution (LTE) network integrated with NDN protocol and our proposed P2C2 replacement policy. The simulation results show that the hitting rate value of P2C2 is higher than Time-aware Least Recently Used (TLRU) and Least Frequently Used (LFU) significantly, and P2C2 reduces the traffic of back-haul while facilitating the offloading of server traffic.