Service provider’s profit maximization in multi-UAV terahertz 6G networks: An AoI and energy-aware multi-agent PPO method

Terahertz (THz) communication is a candidate technology for high-rate data transfer in Sixth-Generation (6G) Internet of Things (IoT) networks under suitable propagation conditions. Data Center (DC) applications require timely updates from energy-constrained IoT sensors to support effective decision-making. Unmanned Aerial Vehicles (UAVs)-assisted Service Provider (SP) offer a viable solution for timely data transfer in exchange for profit, but introduce challenges such as trajectory optimization under strict energy limitations. We formulate the Service Profit Maximization Problem (SPMP), where the SP maximizes profit by optimizing UAV trajectories, with Age of Information (AoI) discounting service revenue and energy consumption determining costs. To solve it, we propose a decentralized sub-optimal approach based on Multi-Agent Deep Reinforcement Learning (MADRL), referred to as Maximizing Profit using Proximal Policy Optimization (MP3O). Simulations under the adopted line-of-sight channel assumptions show that MP3O outperforms the compared methods in profit and energy efficiency.