29 December 2022 in IEEE Transactions on Vehicular Technology
The past few years have witnessed an exponential growth of compelling UAV swarm applications ranging from agricultural production, and intelligent transport, to disaster rescue. The high-speed mobility of the UAV swarm belongs to a new clan of networks, termed flying ad-hoc networks (FANETs). How to design an effective routing mechanism in such a dynamic network is challenging. Traditional flooding searching algorithms (e.g., OLSR, AODV) lead to huge communication overheads, while greedy searching algorithms (e.g., Geolocation-Based Routing protocol) pose low routing efficiency. In this paper, we propose an adaptive communication-based UAV swarm routing algorithm. In our algorithm, we design the Multilayer Perceptron algorithm to learn when routing flooding is required among different UAVs, and design the Gate Recurrent Unit algorithm to greatly compress the volume of communication data. Besides, we further adopt the multi-agent actor-critic algorithm to learn how to integrate shared information for cooperative routing decision-making. Extensive simulation results validate that our algorithms achieve efficient and effective routing under a partially observable distributed environment for large-scale UAV swarm cooperation.
12 March 2024 in IEEE Transactions on Services Computing
The flying ad hoc networks (FANETs) have attracted a large amount of attention from both academia and industry. Benefiting from the flexibility, the FANETs have been widely deployed in various scenarios, ranging from agricultural production to emergency rescue. However, in FANETs, the mobility of unmanned aerial vehicles (UAVs) has led to critical challenges for the stability of communications. Especially, the routing flooding mechanism extremely limits the scalability of FANET. To overcome these technical challenges, constructing a hierarchy and clustering structure in FANETs is considered a promising solution. In this paper, we propose the fission spectral clustering (FSC) strategy for UAV swarm networks. We model the UAV clustering problem as a graph cut problem. The time-sequential attributes weight of nodes and edges will be input to the FSC algorithm. Then, it will construct the Laplace matrix and calculate the first k-th eigenvectors of it. We apply the K-Means algorithm into this feature space to cut the graph by clustering the eigenvectors. Each cluster will constantly fission with this strategy until it satisfies the size and structure constraints in the UAV clusters. Some simulations are implemented to evaluate our proposed algorithm in comparison to the other state-of-the-art solutions.
9 April 2025 in IEEE Transactions on Network Science and Engineering
In recent years, UAVs have been widely employed in the construction of post-disaster emergency communication networks due to their flexible mobility and self-organizing capabilities, enabling them to quickly and spontaneously establish communication links. Post-disaster rescue efforts, restoring connectivity and communication, as well as providing disaster information, all require strong support from network applications. However, an essential element in the development of effective UAV swarm-based applications is the network system. Compared to the fixed networks, UAV swarm networks present unique challenges in design and implementation due to its characteristics of high mobility nodes, unstable links, and dynamic topology. Recently, clustering technology has gained recognition as an effective approach to constructing stable UAV swarm networks. In this paper, for UAV swarm networks, we propose a graph attention-based agglomerative clustering algorithm. This algorithm allows UAV nodes to learn similarity relationships through a graph attention network. By considering the mobility similarity between UAV nodes, each UAV node can merge with its adjacent nodes. Furthermore, we also design a cluster head selection algorithm based on mixed strategy games. The algorithm's effectiveness and accuracy were demonstrated by simulation results, which showed a 80.04% increase in network lifetime compared to baseline algorithms.
16 December 2024 in IEEE Transactions on Mobile Computing
Over the past decade, low-Earth-orbit (LEO) satellite networks have emerged as a critical infrastructure in communication systems, providing wide coverage, high reliability, and global connectivity. Recently, the development of 6G technologies has challenged the LEO satellite networks to guarantee deterministic scheduling for time-sensitive services. However, traditional deterministic networking techniques fall short for LEO satellite networks. Firstly, these techniques impose strict time constraints, but in LEO satellite networks, delay and jitter typically range in the tens of milliseconds, which exceed these limits and render them infeasible. Secondly, the dynamic topologies of LEO satellite networks challenge the inflexible scheduling strategies generated by these techniques, leading to sub-optimal performance and potential strategy failures. To tackle the first problem, we propose a Cycle Specified Queuing and Forwarding (CSQF) based deterministic flows scheduling mechanism. It relaxes strict time constraints by employing cyclic multi-queue scheduling, enabling more flexible and reliable long-distance transmission. For the second problem, we propose a learning-based swarm intelligence method for deterministic flows scheduling in dynamic LEO satellite networks. It includes an algorithm that combines a Dynamic Graph Convolutional Network (DGCN) with an Adaptive Ant Colony Optimization (ACO) algorithm, referred to as the DGCN-ACO algorithm. The DGCN captures the dynamic feature of the network and generates the heuristic information. The Adaptive ACO utilizes the heuristic information and considers each flow's attribute to generate multi-path scheduling strategies for each deterministic flow, as well as updates the DGCN. The experiment results demonstrate the effectiveness of our proposed algorithm.
19 February 2024 in IEEE Transactions on Network Science and Engineering
The rise of Time-Sensitive Networking (TSN) has sparked a significant transformation in the domain of industrial networks. Concurrently, industrial networks have established new benchmarks for communication quality, characterized by stringent requirements for low latency, minimal jitter, and enhanced reliability. In this context, the TSN subcommittee under the IEEE 802.1 working group has released a comprehensive collection of protocol standards. These standards are intricately crafted to ensure deterministic low latency in inherently unpredictable Ethernet environments. To enhance deterministic Quality of Service (QoS) in industrial networks and increase the flexibility of network scheduling paradigms, emerging technologies that combine TSN with Software-Defined Networking (TSSDN) have gained significant attention. Nevertheless, due to the complex and multifaceted nature of factory workflows and the varying scheduling requirements at different hierarchical levels, current industrial scheduling methods encounter challenges in efficiently managing network traffic in real-world factory environments. In this paper, we present a systematic framework designed for the scheduling of factory traffic. Additionally, we introduce two distinct scheduling mechanisms, each based on a framework with relaxed constraints. These mechanisms correspond to intra-level and inter-level traffic transmission in industrial networks. According to our experimental results, we observe an increase of 12.6% to 25.32% in the relaxed constraint scheduling mechanism when compared to the single scheduling scheme, along with a 9% reduction in the proportion of occupied queues.
8 July 2025 in IEEE Transactions on Vehicular Technology
Over recent years, the application of unmanned aerial vehicle (UAV) swarm networks has greatly expanded, encompassing a wide range of areas, including disaster rescue, agricultural production, and intelligent transport. With such rapid expansion, ensuring adaptive resource allocation in UAV swarm networks to meet diverse service demands has become increasingly important. To tackle this challenge, service function chain (SFC) technology has emerged as an enabling solution. Specifically, SFC orchestration aims to provide a flexible and efficient way to embed sequential virtual network functions (VNFs), enabling customized service demands. However, traditional SFC orchestration approaches rely on a centralized controller to generate the orchestration strategy. While the UAV swarm networks are inherently flat and self-organizing, making it impractical to deploy the centralized controller. In this paper, we propose a distributed generative reinforcement learning (GRL) approach for SFC orchestration in highly dynamic UAV swarm networks. It adopts a distributed point-by-point scheme, which avoids global traversing, resulting in lower computational overhead and faster convergence. In this approach, we further propose a distributed GRL-based algorithm, termed the distributed decision transformer actor critic (DDTAC) algorithm. This algorithm combines the decision transformer network with the actor critic network. The decision transformer utilizes its generative learning capability to extract historical local network features and construct the contextual trajectory-based attention feature. Meanwhile, the actor critic framework utilizes this attention feature to adaptively generate the VNFs embedding strategies. Based on the above components, the proposed algorithm effectively captures the highly dynamic features of UAV swarm networks to form a stable SFC orchestration strategy. The experiment findings demonstrate the effective performance of our proposed algorithm.