Speaker: Dr. Atilla Eryilmaz, The Ohio State University
Title: Advances in Adaptive and Proactive Design Methods for the Optimal Control of Data, Social, and Power Networks
Effective control of many complex systems requires the use of adaptive algorithms that track essential system characteristics and react to them in order to achieve the target objective of the particular problem. Such adaptive solutions find use in a wide range of domains including cross-layer resource allocation in communication networks, proactive content distribution in mobile networks, targeting in social networks, protection of infrastructure against adversarial attacks, stable and efficient control of smart power grids, etc. Therefore, development of well-founded approaches and analytical methods for the principled design of adaptive algorithms for the control of complex networked systems is an area of high interest and importance.
Recently, there have been significant advances in the analysis and design of adaptive and, more recently, proactive strategies for the optimal control of stochastic networks. These advances are built on a few key approaches that utilize combinations of optimization and control theories and methods, coupled with novel stochastic analysis and distributed algorithm design techniques. The resulting design methodologies have been proven to be highly versatile and found applications in the control of diverse domains, including those listed above.
In this presentation, I aim to provide an overview of the cornerstones in the recent development of the adaptive and proactive design strategies for stochastic network control. I will provide motivating examples and applications from diverse domains, along with some perspectives on a recurring theme that emerges in these seemingly disparate efforts. I will conclude by discussing some of the ongoing and future efforts that aim to expand the horizon of this framework, both in theory and in applications.
Atilla Eryilmaz is an Associate Professor of Electrical and Computer Engineering at The Ohio State University, where he has been a faculty member since 2007. He received his Ph.D. degree in Electrical and Computer Engineering from the University of Illinois at Urbana-Champaign in 2005. Between 2005 and 2007, he worked as a Postdoctoral Associate at the Laboratory for Information and Decision Systems at the Massachusetts Institute of Technology. He received the NSF-CAREER Award in 2010, and two Lumley Research Awards for Research Excellence in 2010 and 2015, respectively. He is a co-author of the 2012 IEEE WiOpt Conference Best Student Paper, and the 2016 IEEE Infocom Best Paper Awards. He has served as a TPC co-chair of IEEE WiOpt in 2014 and of ACM Mobihoc in 2017, and is an Associate Editor of IEEE/ACM Transactions on Networking since 2015.
Dr. Eryilmaz's research interests lie in the understanding and management of ``complex networked systems'' that naturally arise in a variety of real-world settings, ranging from efficient resource allocation in communication networks, distributed control of smart power grids, the operation of social networks, the management of transportation systems for navigation, etc. In particular, he pursues well-founded development of efficient, robust, adaptable, and scalable algorithms that achieve provably good performance despite the stochastic and highly complex dynamics of the underlying networks.