Organization and Expertise

University of California, Davis (UC Davis):

• Prof. Junshan Zhang:

  Dr. Zhang has a strong background in edge computing, wireless networks, optimization and game theory. Dr. Zhang has been part of the OpenFogConsortium since its very early stages, and has co-founded Smartiply Inc, a Fog/edge computing company to boost network connectivity and embedded real-time edge intelligence for IoT applications. He leads the efforts on Task 1 and is responsible for managing the project.

• Prof. Zhi Ding:

  Dr. Ding has extensive experience in signal processing, coding, and wireless PHY/MAC. His recent works developed efficient CSI feedback and deep-learning for CSI recovery. He leads the efforts on efficient cooperative CSI estimation for distributed learning (Task 5).

Arizona State University (ASU):

• Prof. Gautam Dasarathy:

  Dr. Dasarathy has a strong background in statistical and large scale machine learning. Dr. Dasarathy has had recent experience in designing successful algorithms for ultra-large scale machine learning problems (with billions of features), and will use this experience to design bandlimited sketching algorithms. His recent work demonstrated highly reliable feature selection was possible on ultra-large scale problems. He will lead the efforts on Tasks 2 and 4.

Harvard University:

• Prof. Na Li:

  Dr. Li has extensive experience in distributed learning, optimization and control of networked systems, including theory analysis, algorithm development, and applications. Her recent distributed optimization algorithms were among the fastest distributed gradient descent algorithms. She leads the effort on Task 3.

Collaboration Plan

This project is interdisciplinary in nature, requiring expertise in edge computing, wireless networking, machine learning, signal processing, optimization, and stochastic processes. Some of the PIs have been collaborating successfully for many years. The preliminary work on the distributed zero-order optimization algorithms^[1] were developed by PIs Li and Zhang. PIs Zhang and Ding have recently begun collaboration on deep-learning aided channel estimation. We list some recent publications in the following

References

[1] Y. Tang, J. Zhang, and N. Li, “Distributed zero-order algorithms for nonconvex multi-agent op-timization,” IEEE Transactions on Control of Network Systems. 8 (1) 269 to 281, 2021.

[2] Malu, Mohit and Dasarathy, Gautam and Spanias, Andreas. “Bayesian Optimization in High-Dimensional Spaces: A Brief Survey.” International Conference on Information Intelligence Systems and Applications.

[3] Li, Weizhi and Dasarathy, Gautam and Ramamurthy, Karthikeyan N. and Berisha, Visar. “Finding the Homology of Decision Boundaries with Active Learning.” Advances in neural information processing systems 2020.