Waveguide to Meaning: Semantic-Aware NOMA for Pinching-Antenna Systems

We investigate the performance of the pinching-antenna systems (PASS) for semantic communication (SC) in both single-waveguide and multi-waveguide scenarios, under the constraints of bit-user quality of service (QoS) and bit-to-semantic decoding order in a heterogeneous users downlink non-orthogonal multiple access (NOMA). Multiple pinching antennas in the single-waveguide scenario are at a minimum adjacent spacing required to prevent mutual coupling. An alternating optimization (AO)-based algorithm optimizes users power allocation coefficients and position of pinching antennas in the single-waveguide NOMA framework. For the multi-waveguide scenario, assuming adjacent waveguides at a sufficient lateral distance apart, the waveguides power allocation subproblem is solved using monotonic optimization and minorization-maximization (MM) approach. Specifically, a lower bound surrogate is iteratively maximized under the feasibility constraints such that a non-decreasing sequence of objective is obtained. Numerical results demonstrate that the NOMA based PASS exploiting SC offers higher semantic spectral efficiency (SE) while fulfilling the bit-user QoS requirement when compared to the considered conventional fixed antenna system. Notably, the multi-waveguide scenario becomes more beneficial for creating adjustable wireless channels in stringentconditions with higher bit-user QoS and wider coverage area requirements.