Energy allocation design for the satellite return channel of a massive NOMA system with interference cancellation

The energy allocation problem in a massive access setting under iterative successive interference cancellation (SIC) is undertaken in this work. The classic one-iteration SIC still has room for improvement when the channel coding schemes adopted are non-error free. In this work, we investigate a satellite return link in which transmitters employ the same physical layer and transmission protocol: a given encoder, and spreading-based nonorthogonal multiple access. Leveraging iterative SIC, the gateway is endowed with strong interference suppression potential to deal with many uncoordinated transmissions. We elaborate on the interplay between energy and reliability for all users by designing an energy allocation with minimum energy expenditure when the iterative SIC receiver performs two decoding attempts per user. We resort to the user-asymptotic regime to address the randomness in the multi-feature system model of two-iteration SIC, which has turned the analysis into a deterministic one. We formulate our final design as a vector optimization aided by tools from the calculus of variations. We show the competitive performance achieved by the designed allocation, specially under encoders associated with short-length codes. We also assess our analysis by simulating a low-level SIC implementation. Work supported in part by Agencia Estatal de Investigacion (AEI, Spain) and by European Regional Development Fund (ERDF) through project RODIN (PID2019-105717RB-C22), and in part by the Catalan administration (AGAUR) under Grant 2017 SGR 578 and Fellowship 2020FI B2 00062.