Skip to content
Open access

A Multi-Agent Hybrid Transfer Learning and Monte Carlo Tree Search Framework for Energy Efficient UAV Communication in 6G Aerial Networks

2026 · IEEE Open Journal of the Communications Society · Vol 7, pp. 10997-11009 · 0 citations · 24 references

TL;DR

A lightweight hybrid optimization framework that combines transfer learning and a pre-trained deep neural network with a Monte Carlo Tree Search (MCTS)-based decision-making component to improve the efficiency of multi-agent UAV communication is suggested.

Abstract

Maintaining communication between Uncrewed Aerial Vehicles (UAVs) and Ground Control Stations (GCS) over an extended set of operational distances is still a major challenge in autonomous aerial systems, even within new 6G-enabled aerial network models. The traditional measures of communication management, such as heuristic rule-based approaches and classical deep learning approaches, tend to be inadequate to change the dynamic channel conditions and put computational pressures on resource-constrained UAV platforms. To improve the efficiency of multi-agent UAV communication, this paper will suggest a lightweight hybrid optimization framework that combines transfer learning and a pre-trained deep neural network with a Monte Carlo Tree Search (MCTS)-based decision-making component. The neural network element is used to predict the quality of communication link and the cost of transmitting energy based on real-time signal characteristics of Received Signal Strength Indicator (RSSI), distance of communication, and the level of interference and left over battery power. The MCTS module uses these semantic predictions to dynamically find the best communication strategies, such as direct GCS transmission, relay-assisted routing, and adaptive data-rate control in response to the changing conditions of channels and range. The proposed architecture presented is decentralized, compared to traditional DRL based communication frameworks involving expensive end-to-end policy learning, the proposed TL-MCTS framework divides the lightweight semantic communication estimation from the decentralized search-based optimization to mitigate the computational burden and enhance the adaptability in real-time for large-scale UAV communication environments. Moreover, the framework was tested in dense UAV swarm environment with 10 to 100 UAV agents with repeated experiments and evaluated the communication performance through the help of confidence interval analysis and variance calculation, which validated the communication performance statistically in the dynamic wireless environment.

Read PDF

Similar papers

Open access Oct 2026

AGI-inspired UAV Cluster Communication Self-Organization Framework: Multi-Agent Learning and Network Optimization

Artificial General Intelligence-inspired Multi-Agent Knowledge Transfer (AGI-MAKT) is presented, a novel self-organizing framework for UAV cluster communication networks capable of self-organization and adaptation in complex, resource-constrained environments with limited human supervision.

Fan Zhou · 0 citations
#reinforcement learning Open access Sep 2026

A hierarchical navigation decision-making method for UAV swarms in unknown communication-constrained environments

In recent years, Unmanned Aerial Vehicles (UAVs) have gradually been widely used in various fields such as regional search and disaster relief, and the development of related technologies has also experienced unprecedented growth. Compared to individual UAVs, the collaborative execution of tasks by UAV swarms has more...

Hua-Jie Xiong, Bao-Guo Yu, Jing-Kui Zhang et al. · 0 citations
Open access Aug 2026

Deep Reinforcement Learning–Enhanced Non-Orthogonal Multiple Access Framework for UAV-Assisted Terahertz 6G Communication Networks: A Large-Scale Monte Carlo Simulation Study

Energy-efficiency and outage-probability analyses further confirm the superiority and practical deployability of the proposed DRL-NOMA scheme for next-generation heterogeneous 6G networks.

M. Abdulakreem, Mohanad Mezher · 0 citations
Conference Open access 2026

UAV Path Planning and Multi-Aircraft Collaboration Based on Machine Learning

A three-layer framework of environmental modeling is built -single-machine decision-making, multi-machine collaboration, and systematically sorts out the key technologies of Unmanned Aerial Vehicle path planning-- analytical three-dimensional grid, cylindrical coordinate system spatial modeling, and four-dimensional ri...

Xingran Du, Xi Hong, Min-Rui Li · 0 citations
2026

Cognitive Anti-Jamming for UAV Communications: A Spectrum State Abstraction Empowered Online Learning Approach

This paper studies the anti-jamming channel selection problem for multiple uncrewed aerial vehicle (UAV) communications in dynamic environments with multiple jammers. Considering the distributed and adversarial nature of the scenario, we model the multi-UAV channel selection as a distributed long-term decision-making p...

Xi-Ming Wang, Xin Liu, Zhi-Yong Du et al. · 0 citations
Sep 2026

Self-Attention Enhanced Multi-Agent Reinforcement Learning for UAV Trajectory Design in MEC Networks

Nowadays, Unmanned Aerial Vehicle (UAV)-assisted Mobile Edge Computing (MEC) has been widely recognized as a promising technique for flexibly and effectively handling computation-intensive and delay-sensitive tasks in 5G advanced and 6G networks. This paper investigates the joint optimization of UAV trajectory and comp...

Xue-Wei Zhang, Jian-Tao Li, Yuan Ren et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.