Feducation Series, Artificial Intelligence and Wireless Systems: A Closer Union, by Walid Saad

FIU solid lab’s Federated Education (FeDucation) Webinar Series
Prof. Walid Saad, Virginia Tech
Artificial Intelligence and Wireless Systems: A Closer Union

Related Articles:
[1] M. Chen, Z. Yang, W. Saad, C. Yin, H. V. Poor, and S. Cui, “A Joint Learning and Communications Framework for Federated Learning over Wireless Networks”, IEEE Transactions on Wireless Communications, vol. 20, no. 1, pp. 269 – 283, January 2021. https://arxiv.org/pdf/1909.07972.pdf

[2] M. Kim, W. Saad, M. Mozaffari, and M. Debbah, “On the Tradeoff between Energy, Precision, and Accuracy in Federated Quantized Neural Networks”, in Proc. of the IEEE International Conference on Communications (ICC), Green Communication Systems and Networks Symposium, Seoul, South Korea, May 2022. https://arxiv.org/pdf/2111.07911.pdf

[3] A. T. Z. Kasgari, W. Saad, M. Mozaffari, and H. V. Poor, “Experienced Deep Reinforcement Learning with Generative Adversarial Networks (GANs) for Model-Free Ultra Reliable Low Latency Communication”, IEEE Transactions on Communications, vol. 69, no. 2, pp. 884 – 899, February 2021. https://arxiv.org/pdf/1911.03264.pdf

[4] C. Chaccour, W. Saad, M. Debbah, Z. Han, and H. V. Poor, “Less Data, More Knowledge: Building Next Generation Semantic Communication Networks”, arXiv preprint arXiv:2211.14343, 2022
https://arxiv.org/pdf/2211.14343.pdf

[5] C. Thomas and W. Saad, “Neuro-Symbolic Causal Reasoning Meets Signaling Game for Emergent Semantic Communications”, arXiv:2210.12040, October 2022
https://arxiv.org/abs/2210.12040

Abstract: The last few years witnessed a confluence of two, seemingly disparate research areas: artificial intelligence (AI) and wireless systems. For example, the sixth-generation (6G) of wireless cellular systems will be an AI-native system in which machine learning (ML) techniques will be instilled across the system’s components, protocols, and devices. This confluence led to the emergence of two paradigms at the intersection of AI and communication systems: a) AI for networks (AIN) and b) Networks for AI (NAI). Under the AIN paradigm, the goal is to create data- and knowledge-driven communication protocols and algorithms for wireless system management, operation, and optimization. The design of such AI-native wireless system algorithms requires advances to today’s AI techniques. Meanwhile, the emergence of distributed edge AI frameworks such as federated learning (FL), that require constant wireless communications among heterogeneous learning agents brought forward new networking challenges. In this NAI paradigm, the fundamental question is to understand how system and network constraints, such as wireless channel fading and bandwidth, can impact the accuracy and effectiveness of edge AI and FL algorithms. Indeed, overcoming such system-level questions is a prerequisite for a widespread deployment of FL over real-world wireless systems such as 5G, 6G, and beyond. In this talk, we provide a holistic overview on the current and future state of AI designs for wireless systems. Under the AIN paradigm, we focus on how to create reliable, generalizable ML algorithms for wireless systems within two use cases: the design of reliable, low-latency reinforcement learning for wireless resource management, and the development of reasoning algorithms for an emerging AI-native networking paradigm called semantic communications that can admit many applications of potential interest to Amazon. Then, we turn our attention to the NAI paradigm, and we investigate how, when deployed over real-world wireless networks, the performance of edge AI and FL algorithms will be affected by inherent network properties such as bandwidth limitations, and delay. We also shed some light on how to redesign FL algorithms in a way to take into account device- and system-level resource constraints while striking a tradeoff between accuracy, energy efficiency, and precision. We conclude with an outlook of future research in these two burgeoning areas.

Bio: Walid Saad (S’07, M’10, SM’15, F’19) received his Ph.D degree from the University of Oslo, Norway in 2010. He is currently a Professor at Virginia Tech’s Electrical and Computer Engineering Department where he leads the Network sciEnce, Wireless, and Security (NEWS) group. Dr. Saad is a Fellow of the IEEE. He is also the recipient of the NSF CAREER award in 2013 and the Young Investigator Award from the Office of Naval Research (ONR) in 2015. He was the (co-)author of eleven conference best paper awards at IEEE WiOpt in 2009, ICIMP in 2010, IEEE WCNC in 2012, IEEE PIMRC in 2015, IEEE SmartGridComm in 2015, EuCNC in 2017, IEEE GLOBECOM (2018 and 2020), IFIP NTMS in 2019, IEEE ICC (2020 and 2022). More information: https://www.netsciwis.com/

#federatededucation #wirelesssystems #communication #feducation #distributedlearning #machinelearning #federatedlearning