Video quality and network efficiency trade-offs in video streaming applications

Video Streaming consists of transmitting an encoded video sequence over a network to a receiver that will display the video sequence in real-time. The real-time display requirement means that the video is displayed at the same frame rate as it is captured at the transmitter. For this reason, video data must be delivered by its appropriate deadline.; Applications such as video streaming that allow real-time display of data become increasingly important as the amount of video material available in digital format grows. In this dissertation we study trade-offs of video quality and network efficiency that arise in video streaming applications. The objective is to transmit video of good quality while using network resources efficiently.; We present a framework for evaluating video streaming systems in networks with Quality of Service (QoS) guarantees. In this setting, network cost can be calculated as a function of the network resources required to meet the QoS requirements of the video streaming application. The required network resources can be computed from a traffic descriptor provided by the user during connection setup. The benefit of the received video sequence is quantified by means of a utility function. The objective is then to maximize user surplus. We develop algorithms to select the best combination of traffic descriptor and encoded video sequence.; Wireless networks are becoming an increasingly important component of the modern communication infrastructure. This situation has generated demand for multimedia services in wireless networks. An important consideration in wireless networks is the efficient use of transmitter energy. Efficient use of transmission energy leads to longer battery life of mobile devices, reduced interference to other user and thus greater network capacity. We present a framework where we can study the interaction of the video encoder and a wireless transmitter in a video streaming application. We formulate the problem of transmitting a video sequence using the minimum total transmission energy while meeting the video quality and delay constraints of the streaming application. We consider packetization, mode and quantizer selection at the source coding level jointly with data rate adaptation and packet transmission scheduling at the transmitter.