BEGIN:VCALENDAR VERSION:2.0 PRODID:IEEE vTools.Events//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:Australia/NSW BEGIN:DAYLIGHT DTSTART:20191006T030000 TZOFFSETFROM:+1000 TZOFFSETTO:+1100 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=10 TZNAME:AEDT END:DAYLIGHT BEGIN:STANDARD DTSTART:20190407T020000 TZOFFSETFROM:+1100 TZOFFSETTO:+1000 RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=4 TZNAME:AEST END:STANDARD END:VTIMEZONE BEGIN:VEVENT DTSTAMP:20190716T003214Z UID:A0BACDDD-A4CD-494E-AE26-AF898777BED0 DTSTART;TZID=Australia/NSW:20190715T140000 DTEND;TZID=Australia/NSW:20190715T160000 DESCRIPTION:IEEE Distinguished Lecturer Presentation by\n\nProf. Ekram Hos sain\, Dept. of Electrical & Computer Engineering\, University of Manitoba \, Canada\n\nTitle: Radio Resource Allocation in the Beyond 5G Era: Promis es of Deep Learning and Deep Reinforcement Learning\n\nAbstract:\n\nOptima l resource allocation is a fundamental challenge for dense and heterogeneo us wireless networks with massive wireless connections. Because of the non -convex nature of the optimization problems\, often it is computationally demanding to obtain the optimal resource allocation. Machine learning\, es pecially Deep learning (DL)\, is a powerful tool where a multi-layer neura l network can be trained to model a resource management algorithm using ne twork data. Therefore\, resource allocation decisions can be obtained with out intensive online computations which would be required otherwise for th e solution of resource allocation problems. Recently\, deep reinforcement learning (DRL) has emerged as a promising technique in solving non-convex optimization problems. Unlike deep learning (DL)\, DRL does not require an y optimal/near-optimal training dataset which is either unavailable or com putationally expensive in generating synthetic data. In this talk\, I shal l present a novel centralized DRL-based downlink power allocation scheme f or a multi-cell system intending to maximize the total network throughput. Specifically\, I shall discuss a deep Q-learning (DQL) approach to achiev e near-optimal power allocation policy. I shall present some simulation re sults to compare the proposed DRL-based power allocation scheme with the c onventional schemes in a multi-cell scenario.\n\nBio: Ekram Hossain is a M ember (Class of 2016) of the College of the Royal Society of Canada. Also\ , he is a Fellow of the Canadian Academy of Engineering. Dr. Hossain's cur rent research interests include design\, analysis\, and optimization beyon d 5G/6G cellular wireless networks. He was elevated to an IEEE Fellow “f or contributions to spectrum management and resource allocation in cogniti ve and cellular radio networks”. His research works have received 25\,00 0+ citations (in Google Scholar\, with h-index = 81). He received the 2017 IEEE ComSoc TCGCC (Technical Committee on Green Communications & Computin g) Distinguished Technical Achievement Recognition Award “for outstandin g technical leadership and achievement in green wireless communications an d networking”. Dr. Hossain has won several research awards including the “2017 IEEE Communications Society Best Survey Paper Award” and the “2011 IEEE Communications Society Fred Ellersick Prize Paper Award”. H e was listed as a Clarivate Analytics Highly Cited Researcher in Computer Science in 2017 and 2018. Currently he serves as the Editor-in-Chief of IE EE Press and an Editor for the IEEE Transactions on Mobile Computing. Prev iously\, he served as the Editor-in-Chief for the IEEE Communications Surv eys and Tutorials (2012-2016). He is a Distinguished Lecturer of the IEEE Communications Society and the IEEE Vehicular Technology Society. Also\, h e is an elected member of the Board of Governors of the IEEE Communication s Society for the term 2018-2020.\n\nCo-sponsored by: Abbas Jamalipour\n\n Room: CB11.12.105 Boardroom\, Level 12\, Bldg: Building 11\, University of Technology Sydney \, 81 Broadway\, Ultimo\, NSW 2007\, Sydney\, New Sout h Wales\, Australia\, 2007 LOCATION:Room: CB11.12.105 Boardroom\, Level 12\, Bldg: Building 11\, Unive rsity of Technology Sydney \, 81 Broadway\, Ultimo\, NSW 2007\, Sydney\, New South Wales\, Australia\, 2007 ORGANIZER:a.jamalipour@ieee.org SEQUENCE:2 SUMMARY:Distinguished Lecturer by IEEE VTS NSW Chapter URL;VALUE=URI:https://events.vtools.ieee.org/m/200920 X-ALT-DESC:Description: <br /><p><em><strong>IEEE Distinguished Lecturer Pr esentation</strong> </em><em>&nbsp\;by&nbsp\;</em></p>\n<p>Prof. Ekram Hos sain\, Dept. of Electrical &amp\; Computer Engineering\, University of Man itoba\, Canada</p>\n<p><strong>Title: Radio Resource Allocation in the Bey ond 5G Era: Promises of Deep Learning and Deep Reinforcement Learning</str ong></p>\n<p>&nbsp\;</p>\n<p><strong><em>Abstract:</em></strong></p>\n<p>O ptimal resource allocation is a fundamental challenge for dense and hetero geneous wireless networks with massive wireless connections. Because of th e non-convex nature of the optimization problems\, often it is computation ally demanding to obtain the optimal resource allocation. Machine learning \, especially Deep learning (DL)\, is a powerful tool where a&nbsp\; multi -layer neural network can be trained&nbsp\; to model a resource management algorithm using network data. Therefore\, resource allocation decisions c an be obtained without intensive online computations which would be requir ed otherwise for the solution of resource allocation problems. Recently\, deep reinforcement learning (DRL) has emerged 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 novel centralized DRL-based downlink power allocation scheme for a multi-cell system intending to maximize the total network throughput. Specifically\, I shall discuss a deep Q-learning (DQL) approach to achieve near-optimal power allocation policy. I shall present some simulation results to compare the proposed DRL-based power allocatio n scheme with the conventional schemes in a multi-cell scenario.</p>\n<p>B io:&nbsp\;Ekram Hossain is a Member (Class of 2016) of the College of the Royal Society of Canada. Also\, he is a Fellow of the Canadian Academy of Engineering. Dr. Hossain's current research interests include design\, ana lysis\, and optimization beyond 5G/6G cellular wireless networks. He was e levated to an IEEE Fellow &ldquo\;for contributions to spectrum management and resource allocation in cognitive and cellular radio networks&rdquo\;. His research works have received 25\,000+ citations (in Google Scholar\, with h-index = 81). He received the 2017 IEEE ComSoc TCGCC (Technical Comm ittee on Green Communications &amp\; Computing) Distinguished Technical Ac hievement Recognition Award &ldquo\;for outstanding technical leadership a nd achievement in green wireless communications and networking&rdquo\;. Dr . Hossain has won several research awards including the &ldquo\;2017 IEEE Communications Society Best Survey Paper Award&rdquo\; and the &ldquo\;201 1 IEEE Communications Society Fred Ellersick Prize Paper Award&rdquo\;. He was listed as a Clarivate Analytics Highly Cited Researcher in Computer S cience in 2017 and 2018. Currently he serves as the Editor-in-Chief of IEE E Press and an Editor for the IEEE Transactions on Mobile Computing. Previ ously\, he served as the Editor-in-Chief for the IEEE Communications Surve ys and Tutorials (2012-2016). He is a Distinguished Lecturer of the IEEE C ommunications Society and the IEEE Vehicular Technology Society. Also\, he is an elected member of the Board of Governors of the IEEE Communications Society for the term 2018-2020.</p> END:VEVENT END:VCALENDAR