Faites connaître cette offre !
Reference : UMR8520-FRELEF-040
Workplace : VILLENEUVE D ASCQ
Date of publication : Monday, November 04, 2019
Scientific Responsible name : Pr Laurent CLAVIER
Type of Contract : PhD Student contract / Thesis offer
Contract Period : 36 months
Start date of the thesis : 2 December 2019
Proportion of work : Full time
Remuneration : 2 135,00 € gross monthly
Description of the thesis topic
Project / Subject: Measurement of the quality of the environment and scaling of communicating objects networks.
Innovations / Locks to lift:
- Reliability of communications for IoT (small packets, non-Gaussian interference) and integrity of received data;
- Scaling up and experiments in real environments (confrontation with uses and users);
- Association of regular measurements and sensors reacting to specific events
The smart city, the connected man, the environmental monitoring ... so many applications that need to retrieve information via communicating objects. The number of such objects continues to grow and this is accompanied by a massive increase in wireless transmissions, a heterogeneity of deployed systems and the great variability of environments (in space, in time). Moreover, for energy reasons, the protocols to access the radio resource must be reduced to their minimum (grant free access, Non-Orthogonal Multiple Access - NOMA). A major consequence is the increased impact of interference. Future networks will then face two challenges: robustness and adaptability, with a strong energy constraint and lifespan.
In particular, interference will not present the traditional Gaussian behavior generally assumed in IoT / M2M heterogeneous networks but will have a more impulsive behavior [1,2]. This impulsivity will have a major impact on future networks that it is important to understand, especially for short packets for which one or more strong pulses can drastically degrade performance . This impact remains difficult to determine. If theoretical approaches (such as stochastic geometry for example) attempt to provide answers , the results of experiments are sorely lacking in order to optimize the radio links. Moreover, it is difficult to conceive satisfactory experiments in a laboratory environment since a large number of nodes and a variety of protocols must be deployed and evaluated in an environment that changes over time.
The objectives of the project are to deploy a large number of objects communicating in places of life. We are targeting a specific location, representative of smart cities: Lilliad learning center in Lille University. The first targeted applications concern the quality of the environment (pollution of the area, noise pollution, electro-magnetic pollution) and the management of the learning center (occupancy of places, user movements ...).
The project is structured around two axes:
Axis 1: NOMA solutions for uplink M2M communications. The goal is to define transmission and reception strategies that are both energy efficient and robust in the context of IoT, i.e., for sporadic transmissions and interference that can sometimes behave differently from the traditional Gaussian noise. In particular the design of the receivers and the numerical treatments allowing to implement the non-linearities necessary for the optimal reception will be studied, based on a 28nm FD-SOI technology in conditions of energy efficiency. The gain control and the response of the low noise amplifier can be adapted, the saturations of the amplifier naturally impacting the very nature of the interference. Error correcting codes and MIMO techniques will also be taken into account in the work. Moreover, in the case of LP-WAN approaches, the receiver will have to be able to process a large number of asynchronous signals [5, 6].
Axis 2: Experimentation in a real environment. The objective is to develop the proposed solutions, at least on software defined radio based tools, and to implement it in environments where other technologies are already functional, sometimes in the same frequency bands (at 2.4 GHz for instance).
Assignment to the CSAM group of the IEMN, IRCICA building, CNRS's Haute Borne Campus in Villeneuve d'Ascq
We talk about it on Twitter!