Fluid Antenna-Assisted Edge Gaussian Splatting Using Grey Wolf Algorithm

Edge Gaussian splatting (GS) has emerged as a promising paradigm for delivering high-fidelity immersive experiences on resource-constrained mobile devices by offloading rendering tasks to a proximal edge server. However, its performance is limited by the communication bottleneck. To address this issue, we propose a fluid antenna (FA)-assisted edge GS system. We formulate a joint optimization problem over antenna positions, transmit power, and binary user selection, aiming to minimize a weighted sum of total rendering error and downloading time. This problem is proven to be NP-hard. To solve it efficiently, we develop a grey wolf optimizer (GWO)-based algorithm with a tailored constraint repair mechanism and a greedy decoding strategy to handle non-linear constraints and discrete variables. Simulation results show that the proposed GWO-based algorithm outperforms the fixed antenna position baseline.