Abstract
Unmanned Aerial Vehicle (UAV)-aided two-way relaying systems have attracted widespread attention due to their ability to improve communication efficiency, reduce deployment costs, and enhance reliability. However, most existing systems employ the Time-Division Multiple Access (TDMA) protocol, which suffers from rigid resource allocation and fails to efficiently manage antenna resources within a time slot for multiple users. Furthermore, the reliance on simple Line-of-Sight (LoS) channel models in many studies is often inaccurate, leading to significant performance degradation. To address these issues, this paper investigates a UAV-assisted two-way relaying system based on the Probabilistic Line-of-Sight (PrLoS) model. We propose a novel two-way transmission protocol, termed the Dynamic Dual-Antenna Time-Slot Allocation Protocol (DDATSAP), to facilitate flexible antenna resource allocation for multiple user pairs. To maximize the minimum average message rate for ground users, we jointly optimize the Resource Scheduling Factor (RSF), transmit power, and UAV trajectory. Since the formulated problem is non-convex and challenging to solve directly, we propose an efficient iterative algorithm based on Successive Convex Approximation (SCA) and Block Coordinate Descent (BCD) techniques. Numerical simulation results demonstrate that the proposed scheme exhibits superior performance compared to benchmark systems.