Answers are grounded in this paper and cite their sources.
The rapid growth of computation-intensive and latency-critical applications continues to outpace the hardware evolution of mobile user devices. While Mobile edge computing (MEC) offers a viable solution, traditional terrestrial access points are often constrained by signal blockages and rigid deployment constraints. To address these limitations, this paper proposes a unified optimization framework for an RIS-enabled multi-tier MEC system that synergistically integrates a rotary-wing Aerial Edge Node (AEN) and a Reconfigurable intelligent surface (RIS). Operating under a cross-tier OFDMA protocol, the architecture facilitates a hierarchical computation flow across local devices, a Terrestrial Access Point (TAP), and an AEN. We formulate the problem as a Mixed Integer Non-Convex Program (MINCP) to minimize the total system energy encompassing device transmission, local computation, and edge server processing, subject to strict heterogeneous latency deadlines. To solve this intractable problem, we design a mathematically rigorous Block Coordinate Descent (BCD) algorithm that converges to a Karush-Kuhn-Tucker (KKT) point. The framework systematically (i) optimizes the UAV 3D positioning and RIS passive beamforming via Semidefinite Relaxation (SDR) and Successive Convex Approximation (SCA), (ii) decouples cross-tier user scheduling and power allocation through Big-M constraints and first-order Taylor expansions, and (iii) determines globally optimal partial offloading fractions and CPU resource distributions from a strictly convex cubic subproblem. Numerical results demonstrate that the proposed framework achieves significant energy reductions of up to 68.4% compared to total edge migration and 87.9% compared to standalone local processing. More importantly, compared with the optimization-based Particle Swarm Optimization (PSO) baseline, the proposed structured BCD framework achieves a 31.2% lower energy consumption while providing stable and deterministic convergence. Furthermore, the proposed joint optimization outperforms unoptimized surface reflection and fixed aerial deployment by 55.0% and 64.0%, respectively, while ensuring strict deadline feasibility. Our analysis further uncovers fundamental system insights, including a coherent O(L²) RIS array gain, a convex U-shaped altitude trade-off for the AEN, and the algorithm’s ability to dynamically adapt offloading decisions to avoid edge computational bottlenecks. These findings establish a robust blueprint for the design of energy- and delay-efficient aerial-terrestrial networks in future 6G ecosystems.
No takes yet. Share an insight, caveat, or question.
Alshahrani et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: