A Lagrangian optimization approach to rate control for delay-constrained video transmission over burst-error channels

We propose a rate control algorithm based on Lagrangian optimization for video transmission over burst-error channels. In our rate control approach, the delay and channel capacity constraints in the video transmission are translated into rate constraints at the encoder. Given that a feedback channel is available, the rate control mechanism can dynamically adjust the video encoding rate to meet the changing rate constraints as the channel conditions vary. Lagrangian optimization is used to find the optimal bit-allocation for the input video frames under the rate constraints, with the objective of minimizing the overall distortion at the decoder. We show how the performance of the transmission system, as measured in terms of the received video quality or the data loss rate, can be improved when information about the channel state is available and the encoder has an a priori probabilistic model of the channel behavior.