A Measurement Study on Media Streaming over Wi-Fi in Named Data Networking

Named Data Networking (NDN), aka Content Centric Networking (CCN), excels in content distribution especially multimedia distribution, which can consume significant network bandwidth. With the market penetration of mobile devices and advancement of wireless technologies, media streaming over Wi-Fi becomes increasingly popular but it does not scale well in today's IP based networking. A natural question therefore is how to leverage NDN to improve and optimize media streaming over Wi-Fi. As a first step towards this problem, we set up a 5-node Wi-Fi media streaming test bed based on Wi-Fi Direct technology and use it to collect the bandwidth and CPU usage data when streaming media in NDN as well as in IP networking. We test 4 streaming scenarios in which a live video is streamed from one publisher to multiple consumers over Wi-Fi Direct and present our measurement results in this paper. Our experimental results indicate that the bandwidth consumption between a content publisher and its forwarder (i.e., Access point) over Wi-Fi can be effectively and dramatically reduced by NDN, offering much better scalability than IP. However, CPU usage can become much higher in NDN than in IP, which deserves further investigation and optimization.