Dynamic Placement of Pinching Antennas for Multicast MU-MISO Downlinks

Within the surface-wave communication paradigm, pinching-antenna systems employ dielectric waveguides as both transmission media and reconfigurable radiators, enabling dynamic antenna repositioning to mitigate path-loss limitations inherent in static arrays. This paper integrates pinching-antenna systems into a downlink multiuser multiple-input single-output (MU-MISO) multicast architecture. We first formulate a single-group multicast design problem where we optimize antenna locations along a waveguide under a total power budget to maximize the minimum signal-to-noise ratio (SNR) across users in a max-min fairness sense. Subsequently, we extend this study to a multigroup multicast scenario with multiple waveguides where antenna placement and transmit beamforming are jointly optimized to maximize the minimum signal-to-interference-plus-noise ratio (SINR) across users in all groups. Intergroup interference is eliminated via block-diagonalization beamforming, resulting in a tractable antenna-placement problem. To accommodate different hardware capabilities, we consider both continuous and discrete activation of pinching antennas, yielding four distinct optimization formulations. For each problem, we develop a probability-learning algorithm within a cross-entropy optimization (CEO) framework, iteratively refining a parameterized sampling distribution via closed-form updates. Simulation results demonstrate that all of the proposed algorithms outperform conventional baselines. Notably, the discrete CEO variant achieves performance comparable to its continuous counterpart with significantly lower complexity, offering a favorable performance-complexity trade-off for practical deployment.