Experiences
Research on Delay-Doppler Communications
Supervisor: Assoc. Prof. Shu Sun
Conducting theoretical study and simulation-based research on Orthogonal Time Frequency Space (OTFS) modulation in the delay-Doppler domain for high-mobility wireless communications.
Novel Hybrid Precoder Based on Deep Reinforcement Learning
Supervisor: Assoc. Prof. Qingqing Wu
- Conducted literature review on related research topics and reproduced several relevant GitHub projects.
- Proposed an innovative hybrid precoder architecture that relaxes phase-shifter constraints to optimize S(I)NR.
- Applied (MA)DDPG and SAC to design MU-SISO and MU-MISO hybrid precoders based on deep reinforcement learning.
Multimodal Intelligent Robotic Assistant
Supervisor: Sr. Eng. Pengzhi Chu
- Built a Gradio interface for voice, text, and image input with speech output, context memory, and device control (robotic arm & Micro:bit).
- Used Qwen-Plus to understand instructions and trigger actions such as weather queries, database operations, robot control, and image understanding via Qwen-VL.
- Implemented function calling and MCP in Qwen Agent for tool integration.
Awards
- First Prize (sole award) in the OrangePi Campus Development Competition
IoT-Based Smart Driver Monitoring System
Supervisor: Assoc. Prof. TAN Wee Kek
- Developed a sensor network and processed fatigue-related data using RPi and ESP32 with MQTT for data transmission.
- Built a website for data visualization and video monitoring with WebSocket and MJPG-Streamer.
Awards
- Third Prize (3rd among 12 groups) in the AIoT Group Project
- Individual A+ Performance in the AIoT Course
