Low-altitude Multi-UAV-assisted Data Collection and Semantic Forwarding for Post-Disaster Relief

TL;DR

This work addresses the challenge of data collection and semantic forwarding in post-disaster low-altitude UAV networks by formulating a data collection and semantic forwarding multi-objective optimization problem (DCSFMOP) that jointly optimizes UAV clustering, locations, excitation weights, and semantic symbol counts. It proposes a novel Large Language Model-enabled Alternating Optimization Approach (LLM-AOA) that combines a greedy clustering stage, an LLM-guided NSGA-II for continuous variables, and a greedy optimization for semantic symbols within an AO framework. Results show that LLM-AOA outperforms traditional AOA and other benchmarks, achieving about 26.8% higher transmission rate and 22.9% higher semantic rate under comparable energy, illustrating improved end-to-end throughput and semantic delivery in challenging post-disaster scenarios. The method enables robust, energy-aware data gathering and rapid semantic inference at remote bases, with practical implications for resilient emergency communications and LAE deployment; future work includes solar-powered clustering and distributed heterogeneous UAV semantic networks.

Abstract

The low-altitude economy (LAE) is an emerging economic paradigm which fosters integrated development across multiple fields. As a pivotal component of the LAE, low-altitude uncrewed aerial vehicles (UAVs) can restore communication by serving as aerial relays between the post-disaster areas and remote base stations (BSs). However, conventional approaches face challenges from vulnerable long-distance links between the UAVs and remote BSs, and data bottlenecks arising from massive data volumes and limited onboard UAV resources. In this work, we investigate a low-altitude multi-UAV-assisted data collection and semantic forwarding network, in which multiple UAVs collect data from ground users, form clusters, perform intra-cluster data aggregation with semantic extraction, and then cooperate as virtual antenna array (VAAs) to transmit the extracted semantic information to a remote BS via collaborative beamforming (CB). We formulate a data collection and semantic forwarding multi-objective optimization problem (DCSFMOP) that jointly maximizes both the user and semantic transmission rates while minimizing UAV energy consumption. The formulated DCSFMOP is a mixed-integer nonlinear programming (MINLP) problem that is inherently NP-hard and characterized by dynamically varying decision variable dimensionality. To address these challenges, we propose a large language model-enabled alternating optimization approach (LLM-AOA), which effectively handles the complex search space and variable dimensionality by optimizing different subsets of decision variables through tailored optimization strategies. Simulation results demonstrate that LLM-AOA outperforms AOA by approximately 26.8\% and 22.9\% in transmission rate and semantic rate, respectively.

Figures (7)

• Figure 1: The low-altitude multi-UAV-assisted data collection and semantic forwarding communication network composed of multiple ground users, UAVs, and a remote BS.
• Figure 2: The framework of the proposed LLM-AOA. In the initialization phase, the initial population $\mathcal{P}^{(0)}$ is generated. Then, in each iteration of the AO phase, the UAV clustering assignment $\mathbf{c}$, UAV locations $\mathbf{Q}$ and excitation current weights $\mathbf{w}$, and the number of semantic symbols per UAV cluster $\mathbf{k}$ are sequentially optimized.
• Figure 3: Spatial distribution of UAVs and the corresponding coverage of ground users within a 1000 m × 1000 m monitor area during conducting data collection.
• Figure 4: Solution distributions obtained by the proposed LLM-AOA and other benchmarks.
• Figure 5: Numerical Optimization Results Obtained by the proposed LLM-AOA and other benchmarks.