Constellation Based Symbol Error Rate and Power Constraint Analysis for RSMA

Rate splitting multiple access (RSMA) is a novel multiple access scheme that reduces interference by partially decoding the interference and partially treating the interference as noise. This paper presents a constellation-based analytical framework to derive novel symbol error rate (SER) expressions for a two-user RSMA system by considering both perfect and imperfect decoding of the common stream (CS). The SER expressions are evaluated by considering
$M$
-ary pulse amplitude modulation as users' private stream (PS) and quadrature phase shift keying as a composite CS. Moreover, the power bound analysis is performed to derive power constraints, which solves the constellation ambiguity problem for arbitrary modulation orders of PS. The closed-form average SER (ASER) is then derived under the Rayleigh channel with order statistics. Numerical result demonstrates proper water-falling ASER curves with an increased signal-to-noise ratio and verifies the derived ASER expressions by Monte-carlo simulation. Furthermore, the optimal power assignment is determined that minimizes the overall system's ASER.