In recent years, the need to run machine learning (ML) services over wireless communication networks has promoted the design of new wireless communication protocols capable of efficiently supporting such ML services. In fact, in wireless networks, ML services face major challenges in terms of computation, bandwidth, scalability, privacy, and security. One proposal to overcome such challenges is Over-the-air computation (OAC), which is a known technique where wireless devices transmit values by analog amplitude modulation so that a function of these values (e.g., Federated Learning gradient aggregations) is computed over the communication channel at a common receiver. OAC dramatically reduces communication energy use, amplifies spectrum efficiency of several orders of magnitudes, and achieves privacy protections. The physical reason is the superposition properties of the electromagnetic waves, which naturally return sums of analog values. Consequently, the applications of OAC are almost entirely restricted to analog communication systems. However, the use of digital communications for OAC would have several benefits, such as error correction, synchronization, acquisition of channel state information, and easier adoption by current digital communication systems. Nevertheless, a common belief is that digital modulations are generally unfeasible for computation tasks because the overlapping of digitally modulated signals returns, in general, meaningless values. This talk broke through such belief and presented a fundamentally new computing method, named ChannelComp, for performing OAC by any digital modulation. It also showed how digital modulation formats allow us to compute a broader class of functions than OAC can compute and proposed a feasibility optimization problem that ascertains the optimal digital modulation for computing functions over the air. During the talk, we showed the superior performance of ChannelComp by simulation compared to OAC.
Quick details about the event:
Date: 22 January 2024
Time: 17:00 鈥� 18:00
Venue: Online Event
Click to watch
Speakers鈥� bio:
Dr. Carlo Fischione is a full Professor at KTH Royal Institute of Technology, Electrical Engineering and Computer Science, Division of Network and Systems Engineering (NSE), Stockholm, Sweden. He is Director of the KTH-Ericsson Data Science Micro Degree Program directed to Ericsson globally, Director of the undergraduate education at NSE, Chair of the IEEE Machine Learning for Communications Emerging Technologies Initiative, and founding General Chair of the IEEE International Conference on Machine Learning for Communications and Networking 鈥� IEEE ICMLCN 2024. He is a distinguished lecturer of the IEEE Communication Society.聽 He received the Ph.D. degree in Electrical and Information Engineering (3/3 years) in May 2005 and the Laurea degree in Electronic Engineering (Laurea, Summa cum Laude, 5/5 years) in April 2001, both from the University of L鈥橝quila, Italy. He received the Starting Grant of the Swedish Research Council in 2008. Prof. Fischione has held research positions at Massachusetts Institute of Technology, Cambridge, MA (2015, Visiting Professor); Harvard University, Cambridge, MA (2015, Associate); and University of California at Berkeley, CA (2004-2005, Visiting Scholar, and 2007-2008, Research Associate). He is Honorary Professor at University of L鈥橝quila, Italy, Department of Mathematics, Information Engineering, and Computer Science. His research interests include applied optimization, wireless Internet of Things, and machine learning. He received a number of awards, such as the 鈥淚EEE Communication Society S. O. Rice鈥� award for the best IEEE Transactions on Communications paper of 2018, the best paper award of IEEE Transactions on Industrial Informatics (2007). He is Editor of IEEE Transactions on Communications (Machine Learning for Communications area) and IEEE Transactions on Machine Learning for Communication and Networking, and has served as Associated Editor of IFAC Automatica (2014-2019). He is the co-founder and scientific advisor of ELK. Audio. Prof. Fischione is Member of IEEE (the Institute of Electrical and Electronic Engineers), and Ordinary Member of DASP (the Italian academy of history Deputazione Abruzzese di Storia Patria).聽