Trajectory Design and Beamforming in UAV-Assisted Wireless Networks: A Fine-Tuned M2LLM-Driven DRL-Based Framework
Bibliographic record
Abstract
Optimizing unmanned aerial vehicle (UAV)-assisted wireless networks to serve mobile users (MUs) via beamforming presents significant challenges, mainly due to the dynamic and complex environments. Traditional single-modal data-based modeling methods are often insufficient for capturing the varying environmental characteristics, leading to inaccurate UAV trajectory design and beamforming. To address these issues, we propose a multi-UAV-assisted integrated sensing, communication, and computation (ISCC) framework that processes multi-modal data to enhance environmental awareness and improve communication performance. We then formulate an optimization problem to maximize the average sum rate by jointly optimizing the UAV trajectory and beamforming vectors. Given the non-convex nature of the problem, traditional optimization techniques are inadequate. To this end, we introduce a fine-tuned multi-modal large language model (M2LLM)-driven deep reinforcement learning (DRL)-based joint optimization framework. Specifically, a pre-trained M2LLM is first fine-tuned to predict future MU positions by leveraging historical multi-modal data, including texts, images, and wireless sensing data. The fine-tuned M2LLM is then employed to extract environmental features, where the output of the fine-tuned M2LLM’s last hidden layer is regarded as the environment state vector to eliminate the output uncertainty of the M2LLM. Subsequently, we use a DRL agent to optimize the UAV trajectory and beamforming in a coordinated manner. Extensive simulation results demonstrate that the proposed framework can significantly enhance network performance by enabling environment-aware and adaptive trajectory design and beamforming. The code is available in https://huggingface.co/blYin/MmllmDrlUavTdBf.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".