Linfeng Luo

Linfeng Luo

Auto-generated profile, based on publicly available information

Linfeng Luo

School of Computer Science and engineering, Central South University, Changsha, China

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 May 2025 in IEEE Communications Surveys & Tutorials
Verified
Journal Article
Closed access
ZGZhiqi Guo
Zhiqi Guo
FTFengxiao Tang
Fengxiao Tang
LLLinfeng Luo
Linfeng Luo
MZMing Zhao
Ming Zhao
NKNei Kato
Nei Kato
With the widespread application of digital twin (DT) technology in network optimization and intelligent management, its integration with large language models (LLMs) presents immense potential. LLMs excel in natural language processing, multimodal analysis, and real-time optimization, enabling innovative solutions for intelligent monitoring, resource allocation, and decision-making in complex network environments. This paper systematically reviews the development of DTs and LLMs, elaborates on their core principles and application scenarios, and examines the capabilities of LLM-driven DTs in key network optimization tasks, including traffic prediction, fault diagnosis, resource allocation, and multi-objective optimization. By leveraging real-time data from DTs, LLMs can dynamically generate optimization strategies, enabling precise monitoring and intelligent tuning. Furthermore, this paper explores the potential of integrating LLMs and DTs to address complex challenges such as data quality, latency sensitivity, and energy consumption demands, while summarizing existing technical bottlenecks. Finally, the paper proposes several potential research directions to address these challenges, offering a comprehensive perspective for advancing the efficiency and automation of next-generation intelligent networks.

18 December 2024 in IEEE Transactions on Mobile Computing
Verified
Journal Article
Closed access
FTFengxiao Tang
Fengxiao Tang
LLLinfeng Luo
Linfeng Luo
ZGZhiqi Guo
Zhiqi Guo
YLYangfan Li
Yangfan Li
MZMing Zhao
Ming Zhao
NKNei Kato
Nei Kato
Highly dynamic heterogeneous networks (HDHNs), characterized by high node mobility and heterogeneity, frequently experience complex and recurrent network faults. Conventional centralized fault diagnosis methods demand real-time collection of extensive network-wide data, while distributed approaches often exhibit limited fault detection capabilities. Additionally, machine learning-based fault diagnosis methods are challenged by the scarcity of labeled fault samples required for training. To address these limitations, this study proposes a semi-distributed network fault diagnosis architecture based on a digital twin network (DTN). The proposed architecture facilitates the extraction of a comprehensive labeled fault dataset that closely replicates real-world network conditions. Using this dataset, we perform centralized training of an enhanced anomaly detection model, FTS-LSTM, to infer fault types at the node level. To overcome the drawbacks of both centralized and distributed approaches, we further introduce a semi-distributed fault diagnosis algorithm (SDFD) that integrates fault types and severity levels identified by nodes to infer overall network faults. The proposed fault diagnosis scheme is validated on a semi-physical DTN simulation platform, demonstrating its effectiveness in realistic scenarios.

11 July 2025 in IEEE Wireless Communications
Verified
Journal Article
Closed access
FTFengxiao Tang
Fengxiao Tang
LLLinfeng Luo
Linfeng Luo
ZGZhiqi Guo
Zhiqi Guo
MZMing Zhao
Ming Zhao
NKNei Kato
Nei Kato
As an emerging technology, digital twin (DT) enables real-time system monitoring and decision-making by creating digital replicas of real-world entities. However, existing research typically relies on complex mathematical models, rules, and domain-specific knowledge, overlooking the potential of natural language in extracting and modeling the intrinsic features of entities. In this work, we present the novel concept of semantic twin (ST), utilizing semantic modeling to construct semantic replicas of real-world entities, thereby simplifying the representation of complex systems and enhancing their comprehensibility. Building on the idea of ST, we further extend this concept to network communication systems and propose the semantic twin network (STN) framework. STN employs semantic modeling to map both static and dynamic network data into a unified semantic space, leveraging large language models (LLMs) to drive network reasoning and analysis. This novel network architecture simplifies the modeling and analysis of network systems, offering human-understandable insights that empower non-technical personnel. By leveraging pre-trained LLMs, STN facilitates a deeper network understanding and supports efficient, realtime decision-making in complex and evolving network environments. This article comprehensively discusses the design principles, advantages, and integration of STN with existing technologies, highlighting its broad potential applications in dynamic and complex network environments.

27 February 2026 in IEEE Transactions on Mobile Computing
Verified
Journal Article
Closed access
FTFengxiao Tang
Fengxiao Tang
XWXiaonan Wang
Xiaonan Wang
XYXun Yuan
Xun Yuan
LLLinfeng Luo
Linfeng Luo
MZMing Zhao
Ming Zhao
NKNei Kato
Nei Kato
Network device and system health management is the foundation of modern network operations and maintenance. Traditional health management methods, relying on expert identification or simple rule-based algorithms, struggle to cope with the heterogeneous networks (HNs) environment. Moreover, current state-of-the-art distributed fault diagnosis methods, which utilize specific machine learning techniques, lack multi-scale adaptivity for heterogeneous device information, resulting in unsatisfactory diagnostic accuracy for HNs. In this paper, we develop an LLM-assisted end-to-end intelligent network health management framework. The framework first proposes a multi-scale data scaling method based on unsupervised learning to address the multi-scale data problem in HNs. Secondly, we combine the semantic rule tree with the attention mechanism to propose a Multi-Scale Semanticized Anomaly Detection Model (MSADM) that generates network semantic information while detecting anomalies. Finally, we embed a chain-of-thought-based large-scale language model downstream to adaptively analyze the fault diagnosis results and create an analysis report containing detailed fault information and optimization strategies. We compare our scheme with other fault diagnosis models and demonstrate that it performs well on several metrics of network fault diagnosis.

22 September 2025 in IEEE Transactions on Cognitive Communications and Networking
Verified
Journal Article
Closed access
LLLinfeng Luo
Linfeng Luo
FTFengxiao Tang
Fengxiao Tang
HCHaiyan Chen
Haiyan Chen
SZShuangqian Zhang
Shuangqian Zhang
XCXuehan Chen
Xuehan Chen
MZMing Zhao
Ming Zhao
The increasing heterogeneity of Unmanned Aerial Vehicle (UAV) networks poses significant challenges for coordination, synchronization, and protocol interoperability. While Digital Twin Networks (DTNs) offer a powerful framework for real-time monitoring and control, existing designs fall short in addressing the unique demands of heterogeneous UAV systems. In this paper, we propose Heterogeneous UAV Digital Twin Network (HU-DTN), a novel architecture in which each subnet segment assigns a cluster-head UAV to perform digital twin synchronization and enable cross-segment communication via protocol translation and packet forwarding. This design streamlines communication, reduces bandwidth overhead, and improves synchronization robustness. Based on this architecture, we formulate a cluster-head selection problem aimed at minimizing cross-segment transmission delay under synchronization latency constraints. To solve it, we develop HGQN, a Hypergraph-Q Network that combines deep reinforcement learning with hypergraph neural networks (HGNNs) to model UAV heterogeneity and guide adaptive decisions. Simulation results demonstrate that HU-DTN with HGQN achieves lower delay and higher synchronization success compared to baseline methods.

29 December 2025 in IEEE Transactions on Communications
Verified
Journal Article
Closed access
ZGZhiqi Guo
Zhiqi Guo
FTFengxiao Tang
Fengxiao Tang
LLLinfeng Luo
Linfeng Luo
MZMing Zhao
Ming Zhao
NKNei Kato
Nei Kato
With the rapid development of the internet of things (IoT), IoT devices are now capable of real-time monitoring and collecting environmental and production data through integrated sensors. However, these devices often face challenges related to limited storage capabilities and transmission range. Furthermore, the widespread deployment of IoT devices has raised significant concerns regarding privacy security. To enhance data collection efficiency and ensure the security of location privacy, this study proposes a high altitude platform (HAP) data collection and trajectory design scheme that is aware of location privacy. Firstly, our scheme utilizes HAPs to quickly cover the collection area and transmit data in real time via satellites. Secondly, a differential privacy-based perturbation mechanism is applied to reduce the risk of location information leakage. Finally, the trajectory optimization problem, incorporating privacy awareness, is modeled as a Markov decision process (MDP) and solved using deep reinforcement learning (DRL) techniques to determine the movement decisions of the HAPs. Experimental results demonstrate that this scheme effectively protects location privacy while enhancing the efficiency and security of data collection.