Songyue Han

Dr. Songyue Han

orcid

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

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

Representative Publications

View All Publications
21 December 2024 in Electronics
Verified
Journal Article
Open access
PDF file
SHSongyue Han
Songyue Han
MWMingyu Wang
Mingyu Wang
JZJialong Zhang
Jialong Zhang
SHSongyue Han
Songyue Han
JDJunhong Duan
Junhong Duan
Large language model-related technologies have shown astonishing potential in tasks such as machine translation, text generation, logical reasoning, task planning, and multimodal alignment. Consequently, their applications have continuously expanded from natural language processing to computer vision, scientific computing, and other vertical industry fields. This rapid surge in research work in a short period poses significant challenges for researchers to comprehensively grasp the research dynamics, understand key technologies, and develop applications in the field. To address this, this paper provides a comprehensive review of research on large language models. First, it organizes and reviews the research background and current status, clarifying the definition of large language models in both Chinese and English communities. Second, it analyzes the mainstream infrastructure of large language models and briefly introduces the key technologies and optimization methods that support them. Then, it conducts a detailed review of the intersections between large language models and interdisciplinary technologies such as contrastive learning, knowledge enhancement, retrieval enhancement, hallucination dissolution, recommendation systems, reinforcement learning, multimodal large models, and agents, pointing out valuable research ideas. Finally, it organizes the deployment and industry applications of large language models, identifies the limitations and challenges they face, and provides an outlook on future research directions. Our review paper aims not only to provide systematic research but also to focus on the integration of large language models with interdisciplinary technologies, hoping to provide ideas and inspiration for researchers to carry out industry applications and the secondary development of large language models.

28 February 2022 in Journal of Sensors
Verified
Journal Article
Open access
PDF file
SHSongyue Han
Songyue Han
DMDawei Ma
Dawei Ma
CKChao Kang
Chao Kang
WHWei Huang
Wei Huang
CLChaoying Lin
Chaoying Lin
CTChunyuan Tian
Chunyuan Tian
The development and popularization of mobile Internet and wireless communication technology have spawned a large number of computation-intensive and delay-intensive applications. Limited computing resources and existing technologies cannot meet the performance requirements of new applications. Mobile edge computing technology can use wireless communication technology to offload data to be stored and computing tasks to the nearby assistant or edge server with idle resources. Based on the data offloading of distributed wireless sensor device to device communication, the architecture is designed and the basic framework of distributed mobile edge computing is constructed. To solve the problem of high mobile cloud computing technology, the offloading model of optimized mobile edge computing was proposed, and the stability and convergence of the proposed algorithm were proved. Finally, the system performance of the proposed algorithm is verified by simulation. The results show that the proposed algorithm can converge within a finite number of steps. Compared with other benchmark schemes, the proposed algorithm has better performance in reducing system energy consumption, reducing moving edge response delay and system total delay.

12 September 2025 in Drones
Verified
Journal Article
Open access
PDF file
SHSongyue Han
Songyue Han
PWPengfei Wan
Pengfei Wan
ZLZhixuan Lian
Zhixuan Lian
MWMingyu Wang
Mingyu Wang
DLDongdong Li
Dongdong Li
CFChengli Fan
Chengli Fan
As UAVs are increasingly deployed in complex scenarios such as disaster monitoring, emergency rescue, and power-line inspection, traditional command and control systems face severe challenges in intelligent decision-making, resource allocation, and elastic scalability. To address these issues, we first propose a distributed UAV command and control system based on large language models of the LLaMA2 family. The system adopts a “cloud–edge–terminal” architecture, using 5G as the backbone network and the Internet of Things as a supplement, with edge computing serving as the computing platform. LLMs of various parameter scales are deployed on demand at different hierarchical levels to support both training and inference, enabling intelligent decision-making and optimal resource allocation. Second, we establish a multidimensional system model that integrates computation, communication, and energy consumption, providing a theoretical analysis of network dynamics, resource constraints, and task heterogeneity. Furthermore, we develop an improved Grey Wolf Optimizer (ILGWO) that incorporates adaptive weights, an elite learning strategy, and Lévy flights to solve the multi-objective optimization problem posed by the system. Experimental results show that the proposed system improves task latency, energy efficiency, and resource utilization by 34.2%, 29.6%, and 31.8%, respectively, compared with conventional methods. Real-world field tests demonstrate that, in urban rescue scenarios, the system reduces response latency by 44.7% and increases coordination efficiency by 39.5%. This work offers a reference for the optimized design and practical deployment of UAV command and control systems in complex environments.

28 June 2022 in Computational Intelligence and Neuroscience
Verified
Journal Article
Open access
PDF file
SHSongyue Han
Songyue Han
WHWei Huang
Wei Huang
DMDaWei Ma
DaWei Ma
JGJiLian Guo
JiLian Guo
HHHang He
Hang He
To improve the contradiction between the surge of business demand and the limited resources of MEC, firstly, the cloud, fog, edge, and end collaborative architecture is constructed with the scenario of smart campus, and the optimization model of joint computation offloading and resource allocation is proposed with the objective of minimizing the weighted sum of delay and energy consumption. Second, to improve the convergence of the algorithm and the ability to jump out of the bureau of excellence, chaos theory and adaptive mechanism are introduced, and the update method of teaching and learning optimization (TLBO) algorithm is integrated, and the chaos teaching particle swarm optimization (CTLPSO) algorithm is proposed, and its advantages are verified by comparing with existing improved algorithms. Finally, the offloading success rate advantage is significant when the number of tasks in the model exceeds 50, the system optimization effect is significant when the number of tasks exceeds 60, the model iterates about 100 times to converge to the optimal solution, the proposed architecture can effectively alleviate the problem of limited MEC resources, the proposed algorithm has obvious advantages in convergence, stability, and complexity, and the optimization strategy can improve the offloading success rate and reduce the total system overhead.

17 December 2024 in Drones
Verified
Journal Article
Open access
PDF file
SHSongyue Han
Songyue Han
MWMingyu Wang
Mingyu Wang
JDJunhong Duan
Junhong Duan
JZJialong Zhang
Jialong Zhang
SHSongyue Han
Songyue Han
In emergency rescue scenarios, drones can be equipped with different payloads as needed to aid in tasks such as disaster reconnaissance, situational awareness, communication support, and material assistance. However, rescue missions typically face challenges such as limited reconnaissance boundaries, heterogeneous communication networks, complex data fusion, high task latency, and limited equipment endurance. To address these issues, an unmanned emergency support system tailored for emergency rescue scenarios is designed. This system leverages 5G edge computing technology to provide high-speed and flexible network access along with elastic computing power support, reducing the complexity of data fusion across heterogeneous networks. It supports the control and data transmission of drones through the separation of the control plane and the data plane. Furthermore, by applying the Tammer decomposition method to break down the system optimization problem, the Global Learning Seagull Algorithm for Gaussian Mapping (GLSOAG) is proposed to jointly optimize the system’s energy consumption and latency. Through simulation experiments, the GLSOAG demonstrates significant advantages over the Seagull Optimization Algorithm (SOA), Particle Swarm Optimization (PSO), and Beetle Antennae Search Algorithm (BAS) in terms of convergence speed, optimization accuracy, and stability. The system optimization approach effectively reduces the system’s energy consumption and latency costs. Overall, our work alleviates the pain points faced in rescue scenarios to some extent.

29 January 2026 in Electronics
Verified
Journal Article
Open access
PDF file
KYKe Yang
Ke Yang
SHSongyue Han
Songyue Han
JZJin Zhang
Jin Zhang
YDYan Dou
Yan Dou
GWGang Wang
Gang Wang
Distributed Interactive Simulation (DIS) systems are highly sensitive to temporal delays. Conventional Dead Reckoning (DR) algorithms suffer from limited prediction accuracy and are often inadequate in mitigating simulation latency. To address these issues, a heuristic hybrid prediction model based on Polar Lights Optimization (PLO) is proposed. First, the Transformer architecture is modified by removing the decoder attention layer, and its temporal constraints are optimized to adapt to the one-way dependency of DR time series prediction. Then, a hybrid model integrating the modified Transformer and LSTM is designed, where Transformer captures global motion dependencies, and LSTM models local temporal details. Finally, the PLO algorithm is introduced to optimize the hyperparameters, which enhance global search capability and avoid premature convergence in PSO/GA. Furthermore, a closed-loop mechanism integrating error feedback and parameter updating is established to enhance adaptability. Experimental results for complex aerial target maneuvering scenarios show that the proposed model achieves a trajectory prediction R2 value exceeding 0.95, reduces the Mean Squared Error (MSE) by 42% compared with the results for the traditional Extended Kalman Filter (EKF) model, and decreases the state synchronization frequency among simulation nodes by 67%. This model significantly enhances the prediction accuracy of DR and minimizes simulation latency, providing a new technical solution for improving the temporal consistency of DIS.