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

TL;DR

This paper proposes a multi-UAV network for efficient data collection and semantic forwarding in post-disaster scenarios, optimizing transmission rates and energy use with a novel LLM-enabled optimization approach.

Contribution

It introduces a new multi-UAV data collection and semantic forwarding framework with a large language model-based optimization method for complex, dynamic environments.

Findings

LLM-AOA outperforms traditional methods by 26.8% in transmission rate.

Semantic forwarding improves data efficiency and resilience.

The approach effectively manages complex decision variables in dynamic settings.

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.