Machine Learning Software Development

Li Sun

Ñ San Diego, CA ***** 334-***-**** # ***************@*****.*** ï www.linkedin.com/in/lisun2015 Background and Strengths

• Interdisciplinary Expertise in Computer Science, Applied Mathematics, and Wireless Communication.

• Strong Analytical & Problem-Solving Skills, applying Optimization Theory and Machine Learning to academic and industry challenges, such as KPI forecasting, resource allocation, and traffic steering.

• Solid Technical Proficiency in Machine Learning, with a focus on DL, DRL, RNN, GNN, and Multi-Armed Bandit.

• Extensive Software Development Experience, specializing in Object-Oriented Programming and debugging.

• Hands-on Industry Experience with Docker and Kubernetes for containerized environments. Technical Skills

• Machine Learning & AI: Ray RLlib, PyTorch Lightning, TensorFlow, Keras, NumPy, Pandas, Scikit-learn

• Programming Languages: Python, C/C++, Java, JSP, Delphi

• Development Tools: PyCharm, MATLAB, Visual Studio, Git, LINGO, AMPL, AIMMS, LATEX

• Databases: MongoDB, MySQL, MS SQL Server, Access

• Systems & Frameworks: Linux, macOS, Windows, Docker, Kubernetes, GitHub Project Experience

Fujitsu Network Communications August 2022 – March 2025 Research Scientist Dallas, TX, USA

• Designed a temporal-graph-neural-network-based 5G system performance forecaster enhanced by timestamp encoding. By leveraging GNN and GRU to capture spatial and temporal features of the mobile network topology, along with timestamp context encoding, the model achieves a 46% improvement on prediction accuracy over ARIMA.

• Proposed a multi-agent deep reinforcement learning (PPO)-based intelligent network slicing model for dynamic radio resource management in 5G cellular networks, leading to a 20% performance enhancement.

• Explored fast segment routing strategies for traffic steering in IP networks. Developed both DFS- and BFS-based algorithms to enable rapid traffic steering, efficiently balancing demand flows and alleviating traffic congestion.

• Deployed an O-RAN simulation environment on Linux, built on the Kubernetes platform for orchestrating and scheduling containerized RIC components. The setup allows flexible customization of network topologies and xApps while enabling real-time monitoring of KPIs.

Auburn University October 2016 – May 2022

Graduate Research & Teaching Assistant Auburn, AL, USA

• Designed a space-time contextual handover mechanism based on multi-armed bandit model for 5G mmWave networks. By learning the statistical distribution of users’ mobility, obstacles, and dynamic environment, the approach achieves 30% reduction in handovers in 5G mobility management.

• Developed a distributed access point activation strategy for cell-free massive MIMO networks based on multi-agent deep reinforcement learning. By capturing user demand distribution and inter-cell interference, the method intelligently deactivates redundant APs to conserve energy while maintaining service-level agreement (SLA) guarantees.

• Proposed an SVD-based data compression method to improve precoding efficiency in cell-free massive MIMO cellular networks. By exploiting the physical structure of Rician fading channels, the approach reduces channel matrix transmission overhead by 83.4%.

• Worked as a teaching assistant for courses including Java Programming, Network Security, Computer Architecture, and Operating Systems. Responsibilities included conducting lab sessions, mentoring students, and grading. RWTH Aachen University September 2009 – February 2011 Visiting Scholar Aachen, North Rhine-Westphalia, Germany

• Contributed to a large-scale postal transportation network planning project funded by DHL, focusing on improving cost-efficiency and operational performance using mathematical modeling and optimization techniques.

• Modeled the long-haul shipment scheduling problem in a postal distribution network as a 0-1 integer programming problem and solved it using a dynamic programming algorithm.

• Developed an innovative transportation approach enabling mail vans to exchange shipments en route to reduce costs. Developed a mixed-integer nonlinear programming (MINLP) model to optimize exchange location selection and shipment scheduling. Designed an efficient Tabu Search-based metaheuristic algorithm for solving the problem. Hohai University March 2014 – June 2016

Postdoctoral Fellow Nanjing, Jiangsu, China

• Engaged in modeling and optimization of complex systems in modern port transportation and cold-chain logistics, specializing in intelligent algorithm design.

• Developed a linear programming model using a time-space network framework to effectively capture the dynamics of network flow across spatial and temporal dimensions.

• Designed a customized Tabu Search-based metaheuristic algorithm to solve the problem efficiently, demonstrating superior performance over the LINGO solver in large-scale scenarios.

• Instructed multiple management-related courses, including Supply Chain Management, Marketing Decision Models, and Customer Relationship Management.

Southeast University September 2005 – December 2013 Graduate Research Assistant Nanjing, Jiangsu, China

• Developed information systems for companies across various industries, addressing domain-specific business needs.

• Built a delivery module for a B/S-architecture fresh food logistics management system, capable of generating and visualizing optimal delivery routes for customer orders. The module also supports order consolidation and delivery scheduling decisions.

• Developed a data management module for a C/S-structured pipeline quality information system at Sinopec Engineering Incorporation, supporting local and remote data entry, storage, retrieval, and report generation. Also conducted UI design, module integration, and system testing.

• Led demand analysis, system architecture design, functional specification, and project management for an embedded image-processing system for food contamination tracing. Education

Ph.D. in Computer Science and Software Engineering Auburn University Auburn, AL, USA

Ph.D. & M.S. in Industrial and Systems Engineering Southeast University Nanjing, Jiangsu, China

B.S. in Computer Science and Information Engineering Changzhou Institute of Technology Changzhou, Jiangsu, China Honors and Awards

• Award of INFOCOM Student Conference Grant U.S. National Science Foundation, 2022

• Award of Excellent Master Degree Thesis Academic Degree Committee of Jiangsu Province, 2009

• Award of Excellent Master Degree Thesis Southeast University, 2009

• Second-class Scholarship Southeast University, 2007

• Award of Outstanding Graduate Changzhou Institute of Technology, 2005

• Outstanding Student Scholarship Jiangsu Province, 2004

• First-class Scholarship Changzhou Institute of Technology, 2003

• First-class Scholarship Changzhou Institute of Technology, 2002 Publications (selected)

• Sun L, Hou J, Chapman R. Multi-agent deep reinforcement learning for access point activation strategy in cell-free massive MIMO networks. IEEE INFOCOM - Conference on Computer Communications Workshops, 2023: 1-6.

• Sun L, Hou J, Shu T. Bandwidth-efficient precoding in cell-free massive MIMO networks with rician fading channels. 18th Annual IEEE International Conference on Sensing, Communication, and Networking (SECON), 2021: 1-9.

• Hou J, Sun L, Shu T, et al. The value of traded target information in security games. IEEE/ACM Transactions on Networking, 2021, 29(4): 1853-1866.

• Sun L, Hou J, Shu T. Spatial and temporal contextual multi-armed bandit handovers in ultra-dense mmWave cellular networks. IEEE Transactions on Mobile Computing, 2020, 20(12): 3423-3438.

• Hou J, Sun L, Shu T, et al. Economics of strategic network infrastructure sharing: A backup reservation approach. IEEE/ACM Transactions on Networking, 2020, 29(2): 665-680.

• Sun L, Hou J, Shu T. Optimal handover policy for mmWave cellular networks: A multi-armed bandit approach. IEEE Global Communications Conference (GLOBECOM), 2019: 1-6.

• Sun L, Zhao L, Hou J. Optimization of postal express line network under mixed driving pattern of trucks. Transportation Research Part E: Logistics and Transportation Review, 2015, 77: 147-169.

Contact this candidate