Energy-Efficiency-Oriented Cooperative UAV Caching and Resource Allocation

Abstract

In dynamic hotspot scenarios such as large-scale events and emergency gatherings, UAV-assisted edge content delivery must address challenges like fluctuating user demands, limited cache capacity, and stringent energy constraints. Existing studies primarily focus on delay, throughput, or cache replacement costs, without explicitly considering the coupling between cache and wireless transmission energy. This paper presents an energy-efficiency maximization approach for an edge caching system comprising a remote base station, multiple UAVs, and mobile users. The problem is formulated as a mixed-integer nonlinear program, aiming to maximize the ratio of weighted service utility to total energy consumption. To tackle the computational complexity, we adopt an exponential moving average (EMA) mechanism to track content popularity variations in real time and develop a four-stage alternating optimization algorithm to update user association, cache placement, cooperative transmission, and power allocation. Simulation results demonstrate that the proposed scheme consistently outperforms conventional methods in terms of energy efficiency. Specifically, it improves energy efficiency by $\mathbf{6 1. 3 \%, 4 4. 1 \%}$ and $\mathbf{3 1. 6 \%}$ compared to FIFO, Random and LRU, respectively.