Delay Guarantee and Effective Capacity of Downlink NOMA Fading Channels

• Published in: IEEE journal of selected topics in signal processing 2019-06, Vol.13 (3), p.508-523
• Main Authors: Xiao, Chiyang, Zeng, Jie, Ni, Wei, Liu, Ren Ping, Su, Xin, Wang, Jing
• Format: Article
• Language: English
• Subjects: Algorithms; Channel capacity; Channels; Closed form solutions; Codes; Computer simulation; Delay; delay violation probability; Delays; effective capacity; Exact solutions; Fading; Fading channels; Mathematical analysis; Mellin transforms; NOMA; Nonorthogonal multiple access; Power management; Probability; Quality of service; Resource management; Statistical analysis; stochastic network calculus; Transforms; Upper bounds
• ISSN: 1932-4553; 1941-0484
• DOI: 10.1109/JSTSP.2019.2900938

Nonorthogonal multiple access (NOMA) is promising for increasing connectivity and capacity. But there has been little consideration on the quality of service of NOMA; let alone that in generic fading channels. This paper establishes closed-form upper bounds for the delay violation probability of downlink Nakagami-m and Rician NOMA channels, by exploiting stochastic network calculus (SNC). The key challenge addressed is to derive the Mellin transforms of the service processes in the NOMA fading channels. The transforms are proved to be stable, and incorporated into the SNC to provide the closed-form upper bounds of the delay violation probability. The paper also applies the Mellin transforms to develop the closed-form expressions for the effective capacity of the NOMA fading channels, which measures the channel capacity under statistical delay guarantees. By further applying the min-max and max-min rules, two new power allocation algorithms are proposed to optimize the closed-form expressions, which can provide the NOMA users fairness in terms of delay violation probability and effective capacity. Simulation results substantiate the derived upper bounds of the delay violation probabilities, and the effective capacity. The proposed power allocation algorithms are also numerically validated.