While 5G deployment is being carried out in many places of the world, there has been great interest in the prospects of 5G beyond and the next generation. Among the various visions, a common theme is that artificial intelligence will play a key role, as evidenced by the great interest and advances in machine learning enabled wireless communications and networking. In this talk, we will discuss the motivation, potential, and challenges of incorporating machine learning in wireless communications and networking for 5G and beyond systems.

We will start with two motivating examples, i.e., channel estimation and mobile edge computing, to show why machine learning could be helpful. We will share our experience of several case studies, including (i) a hybrid approach to the classical energy efficiency maximization problem, where traditional models could be used to train a deep learning model; (ii) data augmentation for convolutional neural network (CNN) based automatic modulation classification (AMC), where a conditional generative adversarial network (CGAN) is utilized to generate synthesized training data; and (iii) and an adaptive model for RFID-based 3D human skeleton tracking, which utilizes meta-learning and few-shot fine-tuning to achieve high adaptability to new environments. We will conclude this talk with a discussion of challenges and open problems.

Co-sponsored by: Kingston CA COMSOC Chapter

Speaker(s): Dr. Shiwen Mao,

Agenda:
Virtual Distinguised Lecture by Dr. Shiwen Mao (Auburn University)

6pm (MT) – Introductions

6:10-7:15 – VDL Presentation

7:15-7:30 – Q&A

Denver, Colorado, United States, Virtual: https://events.vtools.ieee.org/m/274419