To support a phenomenal uptick in user demand for video live streaming, the video service providers have to significantly scale up their network resources. To cost-effectively address this problem, we consider the fog computing paradigm, which deploys the computational resource closer to end users for better efficiency and responsiveness. We realize the fog computing features on commercial off-the-shelf devices by implementing 2 schemes. First, we implement FogStream, a scalable live streaming system that enables fog devices, such as routers, to deliver the streaming. Also, we develop Mog, which allows smartphones to share the streaming to its neighbors through Wi-Fi. Experimental study on both schemes shows that they can effectively deliver the live streaming and save the bandw...[ Read more ]

To support a phenomenal uptick in user demand for video live streaming, the video service providers have to significantly scale up their network resources. To cost-effectively address this problem, we consider the fog computing paradigm, which deploys the computational resource closer to end users for better efficiency and responsiveness. We realize the fog computing features on commercial off-the-shelf devices by implementing 2 schemes. First, we implement FogStream, a scalable live streaming system that enables fog devices, such as routers, to deliver the streaming. Also, we develop Mog, which allows smartphones to share the streaming to its neighbors through Wi-Fi. Experimental study on both schemes shows that they can effectively deliver the live streaming and save the bandwidth in realistic scenarios.