With the new wave of digital revolution, wireless communication networks are experiencing a radical paradigm shift from the conventional human-to-human (H2H) communications to machine-to-machine (M2M) communications. To facilitate the massive access of machine-type devices, random access is expected to play a crucial role in the next-generation Internet of Things (IoT) networks.

Thanks to its distributed nature, random access has been widely applied to various wireless networks, including 5G cellular networks and WiFi networks. Yet it has long been observed that the performance of random-access protocols may significantly degrade under heavy traffic. To improve the design of access strategies, reinforcement learning (RL) based approaches have shown tremendous potential. Due to the lack of theoretical guidance, however, they are often designed in an empirical manner. In this talk, I will demonstrate how to leverage our recently proposed unified theoretical framework to optimize the design of RL-based random access. I will conclude the talk by highlighting the challenges and opportunities for intelligent random access for next-generation IoT networks.

Bio-medical and smart sensors applications often require Analog-to-digital converters (ADCs) and Digital-to-analog converters (DACs) with higher than 14 bits resolution, where element mismatch is big issue in high resolution data converter design. The used methods of digital calibration requires a digital post-processing that can be a limit for the resulting overhead. We propose some methods to improve the linearity, and the presented techniques allow directly obtaining the digital output. A chip fabricated in a 180-nm Bipolar-CMOS-DMOS (BCD) process almost obtains the expected SNDR. The experimental result shows that the spurious-free dynamic range (SFDR) improves from 70.71 to 93.29 dB.

Topic:  

1. Construction of Discourse Power for Women in Interdisciplinary Academic Cooperation in the IoT Field and Cultivation of a Gender-Inclusive Culture

2. Advantages and Challenges for Early-Career Young Professionals in Current AI Era