Joint Optimization for UAV-Assisted Mobile Crowdsensing Networks With Heterogeneous Multi-Agent Collaboration
IEEE Transactions on Cognitive Communications and Networking · 2026 · 12 · 8651-8665
DOI: 10.1109/tccn.2026.3699733 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Cognitive Communications and Networking, 2026. Verfasst von Xianyang Deng und 3 weiteren Autor:innen. Die Arbeit nennt 44 Literaturangaben. Gefördert von: Team-based Research Fund, Research Matching, Research Grant Council (RGC) of the Hong Kong Special Administrative Region, China.
Abstract (Originalsprache)
Uncrewed aerial vehicle (UAV)-assisted mobile crowdsensing (MCS) is a new paradigm that leverages user collaboration for data collection. However, limited spectrum resources restrict user participation. Moreover, the heterogeneity among users demands coordination of sensing, communication, and computation. The interdependency among these three phases and resource competition, significantly complicates system optimization. In this paper, we propose a joint optimization framework to tackle the aforementioned challenges. This framework integrates time slot partition for the three phases, resource allocation, and UAV 3D trajectory planning to maximize the processed sensing data volume. To facilitate analysis, we reformulate the problem as a partially observable Markov decision process and solve it with a multi-agent deep reinforcement learning (MADRL) algorithm. However, conventional MADRL based on multi-layer perceptron (MLP) network struggles to capture spatial-temporal dependencies, while also incurring high computational costs and overfitting risks due to excessive parameters. To address this issue, we propose a novel MADRL algorithm with a hybrid actor network. The proposed algorithm integrates convolutional neural networks (CNN) for feature extraction and Kolmogorov-Arnold networks (KAN) to capture state-action dependencies, thereby enhancing training performance across structurally similar time-slot tasks with fewer parameters. Extensive numerical results demonstrate the efficiency of our proposed algorithm compared with various benchmarks.
Förderung
- Team-based Research Fund
- Research Matching
- Research Grant Council (RGC) of the Hong Kong Special Administrative Region, China
- Jiangsu Provincial Key Research and Development Program
- Postgraduate Research & Practice Innovation Program of Jiangsu Province
Zitieren
APA
Deng, X., Liu, W., Fu, Y., & Zhu, Q. (2026). Joint Optimization for UAV-Assisted Mobile Crowdsensing Networks With Heterogeneous Multi-Agent Collaboration. IEEE Transactions on Cognitive Communications and Networking, 12, 8651-8665. https://doi.org/10.1109/tccn.2026.3699733
Vancouver
Deng X, Liu W, Fu Y, Zhu Q. Joint Optimization for UAV-Assisted Mobile Crowdsensing Networks With Heterogeneous Multi-Agent Collaboration. IEEE Transactions on Cognitive Communications and Networking. 2026;12:8651-8665. doi:10.1109/tccn.2026.3699733
BibTeX
@article{Deng2026,
author = {Deng, Xianyang and Liu, Wenshuai and Fu, Yaru and Zhu, Qi},
title = {Joint Optimization for UAV-Assisted Mobile Crowdsensing Networks With Heterogeneous Multi-Agent Collaboration},
journal = {IEEE Transactions on Cognitive Communications and Networking},
year = {2026},
volume = {12},
pages = {8651--8665},
doi = {10.1109/tccn.2026.3699733}
}
EUVIMED