Global Energy Efficiency Maximization in NOMA-Assisted Multi-UAV IoT Networks
Abstract
: In battery-constrained Internet of Things (IoTs) systems, improving global energy efficiency (GEE) is crucial for maximizing network lifetime. We develop a comprehensive optimization framework that encompasses cluster head (CH) selection and rotation, UAV trajectories (UTs), user pairing (UP), and power allocation (PA) using non-orthogonal multiple access (NOMA) technology. This joint optimization problem is nonlinear, nonconvex, computationally intensive, and NP-hard. To address this issue, we employ the block coordinate descent (BCD) method to decompose the joint optimization problem into subproblems and propose a global energy-efficiency maximization algorithm (GEEMA). Initially, we propose a CH selection and rotation framework that uses a weighted CH selection method. Next, UT and UP are jointly optimized through the zebra optimization algorithm (ZOA). Finally, the PA is optimized by calculating a PA factor for each IoT node pair based on their channel gains. The performance of the GEEMA is evaluated against two benchmarks at fixed altitudes of 50 and 100 m, focusing on key parameters such as network scalability, coverage area, power allocation control factors, transmit SNR, and UAV scalability. Simulations demonstrate that GEEMA significantly outperforms benchmark schemes at 40 dB SNR, achieving 78.61% and 20.61% improvements over Benchmarks 1 and 2, respectively, at an altitude of 50 m.