Archives

Abstract: Merging artificial intelligence into the system design has appeared as a new trend in wireless communications areas and has been deemed as one of the 6G technologies. In this talk, we will present how to apply the deep neural network (DNN) for various aspects of physical layer communications design, including the channel estimation, channel prediction, channel feedback, data detection, and beamforming, etc. We will also present a promising new approach that is driven by both the communications data and the communication models. It will be seen that the DNN can be used to enhance the performance of the existing technologies once there is model mismatch. More interestingly, we will show that applying DNN can deal with the conventionally unsolvable problems, thanks to the universal approximation capability of DNN. With the well-defined propagation model in communication areas, we also attempt to explain the DNN under the scenario of channel estimation and reach a strong conclusion that DNN can always provide the asymptotically optimal channel estimations. We have also build test-bed to show the effectiveness of the AI aided wireless communications. In all, DNN is shown to be a very powerful tool for communications and would make the communications protocols more intelligently. Nevertheless, as a new born stuff, one should carefully select suitable scenarios for applying DNN rather than simply spreading it everywhere.

Bio: Prof. Gao’s research interest include signal processing for communications, array signal processing, convex optimizations, and artificial intelligence assisted communications. He has authored/ coauthored more than 150 refereed IEEE journal papers and more than 150 IEEE conference proceeding papers that are cited more than 10000 times in Google Scholar. Prof. Gao has served as an Editor of IEEE Transactions on Wireless Communications, IEEE Journal of Selected Topics in Signal Processing (Lead Guest Editor), IEEE Transactions on Cognitive Communications and Networking, IEEE Signal Processing Letters, IEEE Communications Letters, IEEE Wireless Communications Letters, and China Communications. He has also serves as the symposium co-chair for 2019 IEEE Conference on Communications (ICC), 2018 IEEE Vehicular Technology Conference Spring (VTC), 2015 IEEE Conference on Communications (ICC), 2014 IEEE Global Communications Conference (GLOBECOM), 2014 IEEE Vehicular Technology Conference Fall (VTC), as well as Technical Committee Members for more than 50 IEEE conferences.

Presented by: Dr. Giorgio Quer, Scripps Research 

Date and time: Tuesday, December 15, 2020, from 6 pm to 7:30 pm

Event details and registration info: https://events.vtools.ieee.org/m/248181

Meeting link will be sent at least 2 days in advance of the event to all registrants.

This event is co-organized by IEEE Seattle and Vancouver Joint Communications Chapters.

Abstract: The availability of large longitudinal data opens new opportunities exploiting statistical learning and deep convolutional neural networks in healthcare. We will overview our large study with data collected from wireless devices from 200,000 individuals for 2 years. We will observe how sleep changes significantly among this large population and is related to age and body mass index. We will also show how a person’s resting heart rate may be a uniquely individualized measure of their health, with potential value for early detection of important physiologic changes. This large study provides the baseline for DETECT, our app-based, nationwide clinical study enrolling individuals who routinely use a smartwatch or other wireless devices to determine if individualized tracking of changes in heart rate, activity and sleep can provide early diagnosis and self-monitoring for COVID-19. In this talk, we discuss how this program has been implemented and which insights for the individual and for public health are obtained by analyzing data from more than 36,000 individuals. We show our recent results on the validation of this algorithm, proving that it can identify COVID-19 positive cases by analyzing both self-reported symptoms and wearable sensor data.

Biography: Dr. Giorgio Quer received a Ph.D. degree (2011) in Information Engineering from University of Padova, Italy. In 2007, he was a visiting researcher at the Centre for Wireless Communication at the University of Oulu, Finland. During his Ph.D., he proposed a solution for the distributed compression of wireless sensor networks signals, based on the joint exploitation of Compressive Sensing and Principal Component Analysis. From 2010 to 2016, he was at the Qualcomm Institute, University of California San Diego (UCSD), working on cognitive networks protocols and implementation. At Scripps Research, he is leading the Data Science and Analytics Scripps team involved in the All of Us Research Program (NIH), together with several efforts involving big data and AI in digital medicine, including DETECT, towards the use of wearables to detect COVID-19. He is a Senior Member of the IEEE and a Distinguished Lecturer for the IEEE Communications society. His research interests include wireless sensor networks, probabilistic models, deep convolutional networks, wearable sensors, physiological signal processing, and digital medicine.

Presented by: Mehdi Bennis (U. Oulu) Walid Saad (VirginiaTech), Halim Yanikomeroglu (Carleton U), Wei Yu (U Toronto), and Vincent Wong (UBC)

Date and time: Monday, December 14, 2020, from 8:45 am to 12:15 pm

Event details and registration info: https://events.vtools.ieee.org/m/241334

Meeting link will be sent at least 2 days in advance of the event to all registrants.

This event is co-organized by IEEE Montreal, Ottawa, Turkey, Toronto Young Professionals Affinity Groups, and IEEE Vancouver Joint Communications Chapter.

IEEE Young Professionals Affinity Groups of the Montreal Section, Ottawa Section, Toronto Section, Turkey Section, and the IEEE Vancouver Joint Communications Chapter bring bright minds from the flagship research groups across the globe to give the community technical lectures on cutting-edge areas in wireless communications. This event will cover broad arrays of topics along with fundamental research results targeting a variety of applications in 5G/B5G/6G.

Walid Saad, ECE Department, Virginia Tech, Blacksburg, USA

Professor, Fellow of IEEE

Title: Can Terahertz Communications Provide High-Rate Highly Reliable Low Latency Communications in 6G Networks?

Abstract: Communication at high-frequency terahertz (THz) bands is seen as a staple of the sixth generation (6G) of wireless cellular networks, due to the large amount of available bandwidth. However, 6G systems will have to support, not only high data rates, but also highly reliable communication links for emerging applications such as advanced wireless virtual reality (VR) systems. In particular, advanced wireless VR applications will impose new visual and haptic requirements that are directly linked to the quality-of-experience (QoE) of VR users. These QoE requirements can only be met by wireless 6G connectivity that offers high-rate and high-reliability low latency communications (HRLLC), unlike the low rates usually considered in vanilla 5G ultra-reliable low latency communication scenarios.  Guaranteeing HRLLC in THz-enabled 6G systems requires dealing with the uncertainty that is specific to the THz channel. Therefore, in this talk, after a brief overview on our vision of 6G systems, we will explore the potential of THz for meeting HRLLC requirements. In this regard, we first quantify the risk for an unreliable VR performance through a novel and rigorous characterization of the tail of the end-to-end (E2E) delay. Then, we perform a thorough analysis of the tail-value-at-risk (TVaR)  to concretely characterize the behavior of extreme wireless events crucial to the real-time VR experience. We use this analysis to derive system reliability for scenarios with guaranteed line-of-sight (LoS)  as a function of THz network parameters. We then present simulation results that show how abundant bandwidth and low molecular absorption are necessary to improve the reliability, although their effect remains secondary compared to the availability of LoS, which significantly affects the THz HRLLC performance. We conclude our talk with an overview on other key open problems in the realms of THz communications and 6G systems.

 

Halim Yanikomeroglu, ECE Department, Carleton University, Ottawa, ON, Canada

Professor, Fellow of IEEE, Fellow of Canadian Academy of Engineering, Fellow of Engineering Institute of Canada

 

Title: Wireless Access Architecture: The Next 20+ Years

Abstract: Communication at high-frequency terahertz (THz) bands is seen as a staple of the sixth generation (6G) of wireless cellular networks, due to the large amount of available bandwidth. However, 6G systems will have to support, not only high data rates, but also highly reliable communication links for emerging applications such as advanced wireless virtual reality (VR) systems. In particular, advanced wireless VR applications will impose new visual and haptic requirements that are directly linked to the quality-of-experience (QoE) of VR users. These QoE requirements can only be met by wireless 6G connectivity that offers high-rate and high-reliability low latency communications (HRLLC), unlike the low rates usually considered in vanilla 5G ultra-reliable low latency communication scenarios.  Guaranteeing HRLLC in THz-enabled 6G systems requires dealing with the uncertainty that is specific to the THz channel. Therefore, in this talk, after a brief overview on our vision of 6G systems, we will explore the potential of THz for meeting HRLLC requirements. In this regard, we first quantify the risk for an unreliable VR performance through a novel and rigorous characterization of the tail of the end-to-end (E2E) delay. Then, we perform a thorough analysis of the tail-value-at-risk (TVaR)  to concretely characterize the behavior of extreme wireless events crucial to the real-time VR experience. We use this analysis to derive system reliability for scenarios with guaranteed line-of-sight (LoS)  as a function of THz network parameters. We then present simulation results that show how abundant bandwidth and low molecular absorption are necessary to improve the reliability, although their effect remains secondary compared to the availability of LoS, which significantly affects the THz HRLLC performance. We conclude our talk with an overview on other key open problems in the realms of THz communications and 6G systems.

 

Vincent Wong, ECE Department, University of British Columbia, Vancouver, BC, Canada

Professor, Fellow of IEEE

Title: Throughput Optimization for Grant-Free Multiple Access with Multiagent Deep Reinforcement Learning

Abstract: Grant-free multiple access (GFMA) is a promising paradigm to efficiently support uplink access of Internet of Things (IoT) devices. In this talk, we present a deep reinforcement learning (DRL)-based pilot sequence selection scheme for GFMA systems to mitigate potential pilot sequence collisions. We formulate a pilot sequence selection problem for aggregate throughput maximization in GFMA systems with specific throughput constraints as a Markov decision process (MDP). By exploiting multiagent DRL, we train deep neural networks (DNNs) to learn near-optimal pilot sequence selection policies from the transition history of the underlying MDP without requiring information exchange between the users. While the training process takes advantage of global information, we leverage the technique of factorization to ensure that the policies learned by the DNNs can be executed in a distributed manner. Simulation results show that the proposed scheme can achieve an average aggregate throughput that is close to the optimum, and has a better performance than several heuristic algorithms.

 

Mehdi Bennis, ECE Department, University of Oulu, Finland

Professor, IEEE Fellow

Abstract: This talk will break down the vision of wireless network edge intelligence at scale in terms of theoretical and algorithmic principles in addition to a number of applications in beyond 5G/6G.

 

Wei Yu, ECE Department, University of Toronto, Toronto, ON, Canada

Fellow of IEEE and a Fellow of Canadian Academy of Engineering

Title: Data-Driven Approaches to Wireless Communication System Design

Abstract: In this talk, I will illustrate how machine learning can significantly improve the design of wireless communication systems. I will draw examples from scheduling and power control problems for wireless cellular networks to show that a data-driven approach can circumvent the need for accurate channel estimation and provide near optimal solutions to system-level optimization problems in wireless system design. I will also show how deep neural network (DNN) can be used for efficient and distributed channel estimation, quantization, feedback, and multiuser precoding for massive MIMO systems, thereby providing an efficient solution to a distributed source coding problem. I will conclude by showing the benefit of data-driven design in term of robustness.

 

Title: Oliver Heaviside and Charles Kao

Presented by: Dr. Dan Gelbart, Rapidia

Date and time: Friday, December 11, 2020, from 6 pm to 8 pm

IEEE ComSoc Virtual Distinguished Lecture, co-organized by the IEEE Atlanta, Vancouver, Italy, Windsor, North Macedonia, Egypt, Pune, and Galveston Bay Chapters

Title: Wireless Systems Design in the Beyond 5G Era: Promises of Deep Learning and Deep Reinforcement Learning

Presented by: Prof. Ekram Hossain, University of Manitoba

Date and time: Monday, November 9, 2020, at 8 am

Event details and registration info: https://events.vtools.ieee.org/m/242882

Meeting link will be sent at least 2 days in advance of the event to all registrants.

Abstract: Optimal resource allocation is a fundamental challenge for dense and heterogeneous 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, 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 can be obtained without intensive online computations which would be required 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 supervised DL-based as well as a 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 allocation scheme with the conventional schemes in a multi-cell scenario.

Biography:  Ekram Hossain (F’15) is a Professor in the Department of Electrical and Computer Engineering at University of Manitoba, Canada. He 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, analysis, and optimization beyond 5G/6G cellular wireless networks. He was elevated to an IEEE Fellow “for contributions to spectrum management and resource allocation in cognitive and cellular radio networks”. To date, his research works have received 31,100+ citations (in Google Scholar, with h-index = 91). He received the 2017 IEEE ComSoc TCGCC (Technical Committee on Green Communications & Computing) Distinguished Technical Achievement Recognition Award “for outstanding technical leadership and achievement in green wireless communications and 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”. He was listed as a Clarivate Analytics Highly Cited Researcher in Computer Science in 2017, 2018, and 2019. Currently he serves as the Editor-in-Chief of IEEE Press and an Editor for the IEEE Transactions on Mobile Computing. Previously, he served as the Editor-in-Chief for the IEEE Communications Surveys and Tutorials (2012-2016). He is a Distinguished Lecturer of the IEEE Communications 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.

The session is designed to inspire the audience, to creatively think, innovate and collaborate in defining what would be an early adoption use uses of personalized services that would be a differentiator for smart cities.

Three Takeaways

·      Understand the key drivers and rapid growth and trends of IoT, 5G and AI as disruptive technologies.

·      Power of collaborative technologies in enabling personalized services for smart cities.

·      Examples of use cases leveraging disruptive technologies in key markets health, mobility and public safety for 2020 and beyond.

Fawzi is a senior member of IEEE, Distinguished Lecturer and IEEE ComSoc Regional Director for North America Board (2020-2021) and chairing of  several society chapters (ComSoc, Signal Processing, Computer, EMB, Cons Elec.). Fawzi is the chair for WCNC 2022 and has organized several events such as smart city summit (2019), Blockchain for Healthcare (2018) and Greentech of Smart Cities (2018) and Advanced technologies in Healthcare (2017). Fawzi has been keynote speakers at international conferences and has several publications and co-authored book on “Collaborative Internet of Things for Future Smart Connected Life and Business” published by Wiley.

Fawzi holds a Bachelor of Science with honors and distinction; Masters in Computer Science and Executive MBA.