Mobility-Optimized Dynamic Content Placement for Fast Vehicles in 5G Networks
; Santos, D.
; Figueiredo, Sérgio Figueiredo
; Parreira, B.
Mobility-Optimized Dynamic Content Placement for Fast Vehicles in 5G Networks, Proc IEEE Annual International Symposium on Personal Indoor and Mobile Radio Communications PIMRC, Istambul, Turkey, Vol. , pp. - , September, 2019.
Digital Object Identifier:
Download Full text PDF ( 471 KBs)
This paper presents a framework for improved mobile video delivery in high speed vehicles, integrating Network Function Virtualization, Software Defined Networking and Multi-access Edge Computing (MEC), with user mobility and service consumption context for enabling the dynamic instantiation of video services in edge-positioned virtualized Content Delivery Network (vCDN) nodes. Based on the estimated vehicle (current and future) position, the solution determines: where the video should be dynamically cached and at what time; and which chunks of the video need to be cached at the target location. This "location-aware predictive caching" mechanism is able to operate in a transparent way to both the user and the video service, by handling the necessary flow-based mobility procedures. A proof of concept was implemented and compared with a static caching service and a regular MEC-based vCDN, enabling the assessment of the contributions and impacts of the mechanism. Results show that our solution minimized the number of cache misses in mobility scenarios while reducing backhaul and core network traffic.