Adaptive end-to-end optimization of mobile video streaming using QoS negotiation

Video streaming over wireless links is a non-trivial problem due to the large and frequent changes in the quality of the underlying radio channel combined with latency constraints. We believe that every layer in a mobile system must be prepared to adapt its behavior to its environment. Hence, a layer must be capable of operating in multiple modes; each mode will show a different quality and resource usage. Selecting the right mode of operation requires exchange of information between interacting layers. For example, selecting the best channel coding requires information about the quality of the channel (capacity, bit-error-rate) as well as the requirements (latency, reliability) of the compressed video stream generated by the source encoder. In this paper we study the application of our generic QoS negotiation scheme to a specific configuration for mobile video transmission. We describe the results of experiments studying the overall effectiveness, stability, and dynamics of adaptation of our distributed optimization approach.