Delay Optimal Non-Orthogonal Multiple Access with Joint Scheduling and Superposition Coding.

Xiaoyu Zhao,Wei Chen

DOI: https://doi.org/10.1109/glocom.2017.8254638

2017-01-01

Abstract:As an emergency requirement of Quality of Service (QoS), low latency can be efficiently provided by a cross-layer scheduling. In particular, we consider a pair of Non-Orthogonal Multiple Access (NOMA) users with a random packet arrival over a block fading channel. A probabilistic cross-layer approach is considered to jointly determine the scheduling and superposition coding process, based on the buffer and channel state information. The joint scheduling and coding issue can be formulated as a two-dimension Markov chain, the state of which is a pair of queue lengths, based on which the average delay and power consumption can be obtained. A joint queue and channel aware optimization is formulated to minimize the average delay of packets given an average transmission power constraint. By converting the optimal problem into a linear programming, the optimal delay-power tradeoff can be obtained. We also discover that the optimal policy can be decomposed into a stationary superposition coding and a threshold-based scheduling policy.