Séminaire ICI : Ejder Bastug

Titre du séminaire et orateur

Caching for Computation.

Ejder Bastug, MIT & CentraleSupélec.

Date et lieu

Mardi 21 mars 2017, 11h.

ENSEA, salle 384.


In this talk, we first give a brief overview of edge caching in wireless networks, which is seen as a promising technique for goodput scaling and offloading in 5G (and beyond) cellular networks. We then turn our attention to yet another key phenomenon namely centralized/distributed computing, and show how to perform in-network function computation from information theoretic point of view (functional compression view in particular). Afterwards, we argue how to leverage caching for in-network computation and present a generalized system model of caching for computation. As an application, we shall also introduce a scenario of wireless interconnected virtual reality where caching and computation play an important role, showing that significant immersive experience gains are possible when caching, computation, and computation are handled carefully in various places of the network.

This is a joint work with Mehdi Bennis (University of Oulu, Finland), Mérouane Debbah (CentraleSupélec, France), and Muriel Médard (MIT, USA).


Ejder Bastug is currently a postdoctoral researcher at MIT and CentraleSupélec, working with Muriel Médard and Mérouane Debbah respectively. He obtained his Ph.D. at CentraleSupélec in December 2015, focusing on distributed caching methods in small cell networks. He has been in the executive committee of IEEE WCNC 2014, organizing committee of IEEE BlackSeaCom 2015, technical programme committee of EuCNC 2015, regular TPC member of IEEE conferences, and chair of several international workshops. He is in the organizing committee of one of the two flagship conferences of IEEE Communications Society: IEEE ICC 2017, to be held in Paris. He is a 2016 Exemplary Reviewer of the IEEE Transaction on Communications, representing fewer than 2% of all the reviewers. Moreover, he is the recipient of 2015 of the Supélec Foundation's Best PhD Thesis/Publications Prize, 2016 IEEE Communications Society Best Tutorial Paper Award, 2016 University of Paris-Saclay STIC Doctoral School Best Scientific Contribution Award, Finalist of 2016 Prix Le Monde, and 2017 EURASIP Best Paper Award for the Journal on Wireless Communications and Networking (JWCN). His research interests in the field of information theory, machine learning and wireless communications, with particular focus on functional compression and distributed caching.