| Summary: | 碩士 === 國立臺灣科技大學 === 電子工程系 === 101 === With the rapid evolution of smartphone, the issue of prolonging the battery’s life time has become a major challenge. As the network service becomes popular, the demand of streaming also grows. Many researchers have already proposed works about energy efficient streaming methods through different aspects such as bandwidth control, packet scheduling etc. Those works usually focus on reducing the energy consumption of CPU and wireless network. However, the screen also drains lots of energy, and thus the energy consumption of screen indeed needs to be considered if we want to further extend the battery’s life. In this paper, we proposed an energy efficient streaming system which combined the SVC coding and backlight control mechanism. We used Non-parametric Signal Prediction method to forecast the network condition and then adjust the encoder’s parameters to fit capable network bandwidth. For backlight control, we applied Histogram Equalization to compensate the loss of Image’s contrast owing to reducing the backlight. Experiment results demonstrate that our streaming system can save energy consumption, especially when the user is making a live streaming.
|
|---|
