Current Research (please see publications for details)
Developed an improved transmission waveform compatible with 5G NR standards and 6G candidates and reduced the peak-to-average power ratio (PAPR) to enhance power efficiency and save energy in end-user devices.
在相容於現行5G NR及未來6G可能之波形前提下，設計低峰均功率比(PAPR)之傳輸波形，提升終端裝置功率發射效率，增強省電效果。
Applying the index modulation technique, originally used for multi-antenna systems, to the subcarriers in OFDM systems significantly enhances the trade-off flexibility between spectral efficiency (SE) and energy efficiency (EE) in communication systems.
將原用於多天線系統的索引調變技術運用於正交分頻多工(OFDM)系統中的子載波，大幅增加通訊系統頻譜效益(SE)與能量效益(EE)之權衡彈性。
Utilizing the generalized Papoulis-Gerchberg algorithm and leveraging known system information to compensate for the nonlinear distortion caused by power amplifiers (PA) and impulse noise in wireless communication systems, thereby reducing the bit error rate (BER) of the transmitted signals from IoT devices.
以廣義帕波式迭代演算法(generalized Papoulis-Gerchberg algorithm)利用已知系統資訊來補償無線通訊系統中功率放大器(PA)與脈衝雜訊(impulse noise)所造成之非線性失真，降低物聯網裝置所發送之位元錯誤率(BER)。
Open Radio Access Network (O-RAN) is a rapidly evolving network deployment technology in recent years. By developing the RAN Intelligent Controller (RIC), AI can directly control the parameters of base stations (gNB), enabling automatic optimization of complex B5G networks and enhancing network performance.
開放性無線接入網路 (Open radio access network，O-RAN)為近年急速發展之新式網路佈署技術，透過對RIC (RAN Intelligent Controller)的開發使AI可直接控制基站(gNB)參數，自動優化複雜的B5G網路，提升網路效能。
To support immersive communication scenarios in 6G, fast 3D video rendering technologies based on 3D Gaussian Splatting (3DGS) such as GPS-Gaussian and Gaussian Splatting Coding (GSC) have emerged. Meanwhile, real-time rendering and display technologies on Unity-based end devices are becoming increasingly important.
為滿足6G沉浸式之通訊情境，以3D Gaussian Splatting (3DGS)為基礎之3D視訊快速渲染技術如GPS-Gaussian、GSC (Gaussian Splatting Coding) 因應而生，Unity相關之終端裝置即時顯示技術亦漸趨重要。
In current Intelligent Internet of Things (IoT) architectures, some AI computing tasks traditionally performed in the cloud are offloaded to the perception layer (edge computing). This approach supports privacy protection, low latency, and lightweight, energy-efficient scenarios, while also reducing data transmission and saving network bandwidth.
在現行的智慧物聯網架構中，部分AI應用的雲端運算被下放至感知層(邊緣運算)，用以支援隱私保護、低延遲及輕量省電的情境，並可有效節約數據上傳，節省頻寬。