Lijun Luo

Dr. Lijun Luo


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
5 November 2024 in Symmetry
Verified
Journal Article
Open access
PDF file
KFKan Feng
Kan Feng
LLLijun Luo
Lijun Luo
YXYongjun Xia
Yongjun Xia
BLBin Luo
Bin Luo
XHXingfeng He
Xingfeng He
KPKai Peng
Kai Peng
Large Language Models (LLMs) have demonstrated impressive capabilities in autogenerating code based on natural language instructions provided by humans. We observed that in the microservice models of edge computing, the problem of deployment latency optimization can be transformed into an NP-hard mathematical optimization problem. However, in the real world, deployment strategies at the edge often require immediate updates, while human-engineered code tends to be lagging. To bridge this gap, we innovatively integrated LLMs into the decision-making process for microservice deployment. Initially, we constructed a private Retrieval Augmented Generation (RAG) database containing prior knowledge. Subsequently, we employed meticulously designed step-by-step inductive instructions and used the chain of thought (CoT) technique to enable the LLM to learn, reason, reflect, and regenerate. We decomposed the microservice deployment latency optimization problem into a collection of granular sub-problems (described in natural language), progressively providing instructions to the fine-tuned LLM to generate corresponding code blocks. The generated code blocks underwent integration and consistency assessment. Additionally, we prompted the LLM to generate code without the use of the RAG database for comparative analysis. We executed the aforementioned code and comparison algorithm under identical operational environments and simulation parameters, conducting rigorous result analysis. Our fine-tuned model significantly reduced latencies by 22.8% in handling surges in request flows, 37.8% in managing complex microservice types, and 39.5% in processing increased network nodes compared to traditional algorithms. Moreover, our approach demonstrated marked improvements in latency performance over LLMs not utilizing RAG technology and reinforcement learning algorithms reported in other literature. The use of LLMs also highlights the concept of symmetry, as the symmetrical structure of input-output relationships in microservice deployment models aligns with the LLM’s inherent ability to process and generate balanced and optimized code. Symmetry in this context allows for more efficient resource allocation and reduces redundant operations, further enhancing the model’s effectiveness. We believe that LLMs hold substantial potential in optimizing microservice deployment models.

2 January 2025 in Symmetry
Verified
Journal Article
Open access
PDF file
GWGeng Wu
Geng Wu
LJLu Jing
Lu Jing
DHDai Hou
Dai Hou
LZLei Zhen
Lei Zhen
KLKaihong Li
Kaihong Li
KPKai Peng
Kai Peng
In recent years, unmanned aerial vehicles (UAVs, also known as drones) have gained widespread application in fields such as data collection and inspection, owing to their lightweight design and high mobility. However, due to limitations in battery life, UAVs are often unable to independently complete large-scale data collection tasks. To address this limitation, vehicle–drone collaborative data collection has emerged as an effective solution. Existing research, however, primarily focuses on collaborative work in static task scenarios, overlooking the complexities of dynamic environments. In dynamic scenarios, tasks may arrive during the execution of both the vehicle and UAV, and each drone has different positions and remaining endurance, creating an asymmetric state. This introduces new challenges for path planning. To tackle this challenge, we propose a 0–1 integer programming model aimed at minimizing the total task completion time. Additionally, we introduce an efficient dynamic solving algorithm, referred to as Greedy and Adaptive Memory Process-based Dynamic Algorithm (GAMPDA). This algorithm first generates an initial global data collection plan based on the initial task nodes and dynamically adjusts the current data collection scheme using a greedy approach as new task nodes arrive during execution. Through comparative experiments, it was demonstrated that GAMPDA outperforms SCAN and LKH in terms of time cost, vehicle travel distance, and drone flight distance and approaches the ideal results. GAMPDA significantly enhances task completion efficiency in dynamic scenarios, providing an effective solution for collaborative data collection tasks in such environments.

23 February 2026 in Applied Sciences
Verified
Journal Article
Open access
PDF file
JLJieke Lu
Jieke Lu
XYXinyi Yang
Xinyi Yang
YLYang Liu
Yang Liu
HZHaoran Zuo
Haoran Zuo
FZFeng Zhou
Feng Zhou
LLLijun Luo
Lijun Luo
With the rapid growth of large-scale networks, traditional rule-based and supervised anomaly detection methods struggle with heavy reliance on labeled data, slow response to rapidly changing patterns, and difficulty in capturing complex temporal anomalies. At the same time, real-world traffic exhibits strong class imbalance, where normal samples overwhelmingly dominate, causing many existing models to miss subtle but critical abnormal behaviors. To address these challenges, this paper proposes an unsupervised temporal anomaly detection framework for network traffic based on a Transformer-autoencoder bidirectional prediction and reconstruction model. The framework combines the advantages of autoencoders and regression models, using multi-head self-attention and positional encoding to capture long-range temporal dependencies in traffic sequences. A masked decoding mechanism is further employed to prevent information leakage from future time steps. The model jointly generates forward and backward predictions as well as reconstructed sequences, and designs multiple anomaly scoring strategies that integrate prediction and reconstruction errors to enhance the sensitivity to point, contextual, and collective anomalies under highly imbalanced data. Experiments on three public benchmark datasets demonstrate that the proposed method significantly improves detection performance, achieving up to an F1 score of 0.960 and a precision of 0.949, with recall approaching 1.0, while reducing false alarms, thereby showing strong applicability to practical network security scenarios.