Mobile cloud gaming has recently gained popularity as a result of improvements in the quality of internet connections and mobile networks. Under stable conditions, current LTE networks can provide a suitable platform for the demanding requirements of mobile cloud gaming. However, since the quality of mobile network connections constantly change, the network may be unable to always provide the best possible service to all clients. Thus, the ability to adapt is necessary for a mobile cloud gaming platform in order to compensate for changing bandwidth conditions in mobile networks. One approach for doing this is to change the quality of the video stream to match the available bandwidth of the network.This thesis evaluates an adaptive streaming method implemented on a mobile cloud gaming platform called GamingAnywhere and provides an alternative approach for estimating the available bandwidth by measuring the signal strength values of a mobile device. Experimentation was conducted in a real LTE network to determine the best approach in reconfiguring the encoder of the video stream to match the bandwidth of the network. The results show that increasing the constant-rate-factor parameter of the video encoder by 12 reduces the necessary bandwidth to about half. Thus, changing this video encoder parameter provides an effective means to compensate for significant changes in the bandwidth. However, high values of the constant-rate-factor parameter can considerably reduce the quality of the video stream. Thus, the frame rate of the video should be lowered if the constant-rate-factor already has a high value.
展开▼
