🔬 Research Experience</h2>
Research Intern | Tsinghua University Lab
Mar 2025 – Present | Beijing, China
- Conducting research on intelligent mobility management in high-speed cellular networks (e.g., metro scenarios), aiming to enhance streaming experience by predicting network buffering events.
- Designed a cross-layer stalling prediction model using physical-layer signals (RSRP, SNR, PCI) and higher-layer metrics; achieved an F1-score of ~0.68 through feature selection and iterative model training.
- Performed real-world data analysis using CDF plots to identify signal fluctuation patterns during handovers, enabling proactive mitigation strategies.
- Deployed the model prototype on mobile devices, demonstrating reduced stalling frequency and duration compared to baseline approaches.
Core Member | Edge-Cloud Collaborative Deep Learning Inference Framework
Dec 2023 – Dec 2024 | Chongqing University (Chongqing Municipal Innovation Program)
- Developed a dynamic model partitioning framework to reduce computation latency and energy consumption on edge devices by offloading partial inference to edge servers.
- Implemented and evaluated the system on diverse DNN architectures including AlexNet, ResNet, and YOLO, achieving improved throughput and lower end-to-end latency.
- Compared fixed vs. dynamic partitioning strategies; results showed significant speedup over pure edge-side inference.
- Contributed to a manuscript currently under preparation for submission.
Project Lead | Sentiment Analysis & Prediction on Distributed Clusters
Apr 2024 – Jul 2024 | Course Design Project, Chongqing University
- Led a team to build a data processing pipeline using Hadoop and Spark for large-scale text sentiment analysis.
- Applied traditional machine learning models (e.g., SVM, Random Forest) and fine-tuned BERT-based neural networks for sentiment classification and trend prediction.
- Gained hands-on experience in full-cycle software development: requirement analysis, prototyping, implementation, and deployment.
Project Lead | Traffic Flow Prediction using Graph Neural Networks
Oct 2022 – Oct 2023 | SRTP Project, Chongqing University
- Led a team to model urban traffic networks using Graph Convolutional Networks (GCN) for short-term traffic flow forecasting.
- Reproduced and analyzed state-of-the-art models, gaining foundational skills in academic literature review, environment setup, and code reproduction.
- Explored node-level prediction tasks on real-world road network datasets.
Independent Project | “XianYu” Campus Second-hand Trading App
2023 | Full-stack Development
- Designed and developed a mobile application prototype for university students to trade second-hand goods.
- Practiced full software development lifecycle: requirements gathering, UI/UX prototyping, coding, and deployment.
Hardware Project | CPU Design from Scratch using Verilog
2023 | Computer Architecture Course
- Designed a complete CPU from scratch using Verilog, including control unit, data path, and 52 basic/privileged instructions.
- Implemented a Cache module to optimize memory access speed and reduce latency.