Near-optimal delay-constrained MLR scheduler for multimedia embedded wireless channels

We consider real-time video packet transmission over a wireless channel. In our model, optimal rate allocation strategies are determined by the cross-layer perspective that minimize the average power consumed by the transmitter under the average delay constraints. We present simple packet scheduling policies which are influenced by both the queue and the channel state and those can be utilized to tradeoff characteristic between queueing delay and transmission power. The near-optimal scheduler adopts a low-complexity modified-linear-rule (MLR) which tends to make the queue evolve as a stable feed-back linear system. The parameter is chosen so as to reach the desired compromise between delay and power consumption. Throughout the extensive simulations, we generalize the scheme of MLR scheduler even in the multiuser wireless fading environments, compare it with the performance of the optimal scheduler and the log-linear scheduler and derive the optimal rate adaptation for video sources under the average delay and power constraints.