> Academic opportunities > PHD positions

Power Efficient AI-based IoT Physical Layer

Technological challenge: Communication networks, IOT, radiofrequencies and antennas (learn more)

Department: Département Systèmes (LETI)

Laboratory: Laboratoire Communication des Objets Intelligents

Start Date: 01-09-2022

Location: Grenoble

CEA Code: SL-DRT-22-0712

Contact: valerian.mannoni@cea.fr

With the proliferation of connected objects, the Internet of Things (IoT) is currently at the center of digital evolution. It is now possible to connect daily / industrial objects (robots, cars, sensors, etc.) to the Internet via embedded terminals. The mMTC (massive Machine Type Communications) connectivity technologies of 5G, have only slightly evolved from previous generations and are not effective in meeting the IoT challenges of energy efficiency, coverage and complexity. Thus, it becomes necessary to evolve IoT waveforms for future generations of telecommunication systems (6G). In this context, we want to propose and study a new physical layer that meets the main IoT requirements which are a good energy efficiency (battery operation), a long communication coverage and this, for mainly uplink sporadic communications with small packets. To do so, we will take advantage of the progress made in deep learning and in particular of the auto-encoder architectures. Thus, after having proposed and evaluated a first trainable IoT physical layer, the synchronization and channel estimation processes will be integrated. The agility of the resulting physical layer to changes in its environment (propagation channel, multiple access, new service requirements, etc.) will then be studied as well as its implementation on embedded AI architectures.

See all positions Download the offer (.zip)

Email Bookmark and Share