Experimental assessment of Big Data-backed video distribution in the telecom cloud

The telecom infrastructure is undergoing a huge transformation since operators are deploying their own cloud infrastructure to provide cloud services and enabling Network Functions Virtualization (NFV). The resulting infrastructure is referred to as the telecom cloud. NFV decouples network functions from proprietary hardware appliances, so they can be implemented in software and deployed on virtual machines (VM) running on commercial off-the-shelf computing hardware. One of the network functions currently being considered by many operators is Content Delivery Network (CDN) for live-TV and Video on Demand (VoD) distribution. In fact, a significant advance in video delivery is the standardized MPEG Dynamic Adaptive Streaming over HTTP (MPEG-DASH) technique that enables media content delivering over the Internet using standard HTTP web server infrastructure. Besides, the CDN reconfiguration, motivated by an increase in the number of users demanding content, entails a Virtual Network Topology (VNT) reconfiguration to cope with the demanded connection capacities. A Big Data-based CDN manager, fulfilling the ETSI NFV guidelines, is proposed to adapt the virtualized CDN function to current and future demand. Its feasibility to control virtualized components while collecting monitoring data is experimentally assessed in a real environment.