BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:US/Eastern
BEGIN:DAYLIGHT
DTSTART:20210314T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20201101T010000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20201111T004443Z
UID:1558E1D2-74A5-42C5-B302-8DF0C79BE5AB
DTSTART;TZID=US/Eastern:20201109T110000
DTEND;TZID=US/Eastern:20201109T121500
DESCRIPTION:Just 2 days left to register!!\n\nOptimal resource allocation i
 s a fundamental challenge for dense and heterogeneous wireless networks wi
 th massive wireless connections. Because of the non-convex nature of the o
 ptimization problems\, often it is computationally demanding to obtain the
  optimal resource allocation. Machine learning\, especially Deep learning 
 (DL)\, is a powerful tool where a multi-layer neural network can be traine
 d to model a resource management algorithm using network data. Therefore\,
  resource allocation decisions can be obtained without intensive online co
 mputations which would be required otherwise for the solution of resource 
 allocation problems. Recently\, deep reinforcement learning (DRL) has emer
 ged as a promising technique in solving non-convex optimization problems. 
 Unlike deep learning (DL)\, DRL does not require any optimal/near-optimal 
 training dataset which is either unavailable or computationally expensive 
 in generating synthetic data. In this talk\, I shall present a supervised 
 DL-based as well as a centralized DRL-based downlink power allocation sche
 me for a multi-cell system intending to maximize the total network through
 put. Specifically\, I shall discuss a deep Q-learning (DQL) approach to ac
 hieve near-optimal power allocation policy. I shall present some simulatio
 n results to compare the proposed DRL-based power allocation scheme with t
 he conventional schemes in a multi-cell scenario.\n\nSpeaker(s): Dr. Ekram
  Hossain\, \n\nAtlanta\, Georgia\, United States\, Virtual: https://events
 .vtools.ieee.org/m/242882
LOCATION:Atlanta\, Georgia\, United States\, Virtual: https://events.vtools
 .ieee.org/m/242882
ORGANIZER:syed.tamseel@ieee.org
SEQUENCE:7
SUMMARY:Virtual Dist. Lecture-Wireless Systems Design in the Beyond 5G Era:
  Promises of Deep Learning and Deep Reinforcement Learning
URL;VALUE=URI:https://events.vtools.ieee.org/m/242882
X-ALT-DESC:Description: <br /><p><span style="font-size: 14pt\; color: #ff0
 000\;"><strong>Just 2 days left to register!!</strong></span></p>\n<p>&nbs
 p\;</p>\n<p>Optimal resource allocation is a fundamental challenge for den
 se and heterogeneous wireless networks with massive wireless connections. 
 Because of the non-convex nature of the optimization problems\, often it i
 s computationally demanding to obtain the optimal resource allocation. Mac
 hine learning\, especially Deep learning (DL)\, is a powerful tool where a
  multi-layer neural network can be trained to model a resource management 
 algorithm using network data. Therefore\, resource allocation decisions ca
 n be obtained without intensive online computations which would be require
 d otherwise for the solution of resource allocation problems. Recently\, d
 eep reinforcement learning (DRL) has emerged as a promising technique in s
 olving non-convex optimization problems. Unlike deep learning (DL)\, DRL d
 oes not require any optimal/near-optimal training dataset which is either 
 unavailable or computationally expensive in generating synthetic data. In 
 this talk\, I shall present a supervised DL-based as well as a centralized
  DRL-based downlink power allocation scheme for a multi-cell system intend
 ing to maximize the total network throughput. Specifically\, I shall discu
 ss a deep Q-learning (DQL) approach to achieve near-optimal power allocati
 on policy. I shall present some simulation results to compare the proposed
  DRL-based power allocation scheme with the conventional schemes in a mult
 i-cell scenario.</p>
END:VEVENT
END:VCALENDAR