Stochastic Dynamic Programming for Network Resource Allocation

#Stochastic #Dynamic #Programming #(SDP) #Markov #Decision #Processes #Partially #Observable #MDP #(POMDP) #Network #Resource #Allocation
Share

Stochastic Dynamic Programming (SDP) is a powerful tool for Markov Decision Processes and Partially Observable MDP (POMDP), etc. However, the commonly used Bellman equation and its corresponding backwards recursion algorithm suffer from prohibitive computational complexity because of the curse of dimensionality.   This talk presents an application of the SDP method to wireless network resource allocation and develops a 3-layer decomposition algorithm to achieve global optimum solution with affordable complexity.  Numerical results show that the SDP algorithm achieves better look-ahead planning than the greedy algorithm and other heuristic algorithms. Practical applications of the research include radio spectrum sharing, intelligent transportation systems, and underwater wireless sensor networks for structure health monitoring systems.  


  Date and Time

  Location

  Hosts

  Registration


  • Date: 19 Sep 2019
  • Time: 02:00 PM to 03:00 PM
  • All times are (GMT-05:00) Canada/Eastern
  • Add_To_Calendar_icon Add Event to Calendar
  • 87 Gerrard Street
  • Toronto, Ontario
  • Canada M5B 2K3
  • Building: EPH (Eric Palin Hall)
  • Room Number: 147
  • Click here for Map
  • Contact Event Host

  Speakers

Topic:

Stochastic Dynamic Programming for Network Resource Allocation

Stochastic Dynamic Programming (SDP) is a powerful tool for Markov Decision Processes and Partially Observable MDP (POMDP), etc. However, the commonly used Bellman equation and its corresponding backwards recursion algorithm suffer from prohibitive computational complexity because of the curse of dimensionality.   This talk presents an application of the SDP method to wireless network resource allocation and develops a 3-layer decomposition algorithm to achieve global optimum solution with affordable complexity.  Numerical results show that the SDP algorithm achieves better look-ahead planning than the greedy algorithm and other heuristic algorithms. Practical applications of the research include radio spectrum sharing, intelligent transportation systems, and underwater wireless sensor networks for structure health monitoring systems.  

Biography:

Yahong Rosa Zheng received the Ph.D. degree Carleton University, Ottawa, ONT, Canada, in 2002. From 2003 to 2005, she was an NSERC Postdoctoral Fellow with the University of Missouri-Columbia. From 2005 -2018, she was on the faculty of the Department of Electrical and Computer Engineering at the Missouri University of Science and Technology where she held the Wilkens' Missouri Telecommunications Endowed Professor position for 2017 - 2018. She joined Lehigh University in Aug. 2018 as a professor in the ECE department. Her research interests include underwater cyber-physical systems, real-time embedded systems and signal processing, wireless communications, and wireless sensor networks. She has served as a Technical Program Committee (TPC) member for many IEEE international conferences. She served as an Associate Editor for IEEE Transactions on Wireless Communications 2006-2008 and IEEE Transactions on Vehicular Technology 2008-2016. She is currently Associate Editor for the IEEE Journal of Oceanic Engineering since 2016. She is the recipient of an NSF faculty CAREER award in 2009. She has been an IEEE fellow and a Distinguished Lecturer of IEEE Vehicular Technology Society since 2015.

Email: