Friday 12 January 2018
Contract: for AT MOST 10 months
Start date: ASAP
Location: Grenoble, Rhône-Alpes, France
Hosting institution: LIG and VERIMAG laboratories, Université Grenoble Alpes, Grenoble Institute of Technology
Scientific advisors: Olivier Alphand, Franck Rousseau @LIG/UGA, Karine Altisen, Stéphane Devismes @VERIMAG/UGA
Application deadline: 01/03/2018
How to Apply & Contact Information
Please send the following documents to Franck.Rousseau@imag.fr and Karine.Altisen@univ-grenoble-alpes.fr
Email subject MUST start with "[Post-Doc IoT]"
Letter of application: why you are interested in this multidisciplinary research position and what you would like to work on?
References or letters of recommendation
Applicant’s scientific report or paper written in English
Any other document showing that you are an outstanding candidate
Standardization is undergoing a fast pace at all the layers of the protocol stack. In particular, the IEEE 802.15.4 standards are dedicated to the physical and link layers, where Medium Access Control (MAC) is of primary interest. On top of them, 6LowPAN/IPv6 and RPL, standardized at the IETF, are dealing with the network layer and routing. IEEE 802.15.4e defines TSCH (IEEE Std 802.15.4e-2012), a new MAC which focuses on ultra low-power and high reliability, targeting industrial applications. TSCH is named upon its two key principles: Time Slotted operations and Channel Hopping. The former reduces the radio activity to the minimum, providing ultra low-power operation, while both together guarantee time and frequency diversity, hence providing high reliability. More precisely, TSCH organizes communications using \em cell reservations, where each cell is defined as a pair made of a time slot and a channel.
Now, TSCH defines only how to execute a communication schedule. For this reason, the IETF is currently working on the standardization of 6TSCH, an intermediate layer that will provide the missing parts. In this scope, support for resource allocation is being defined and will be compatible with both centralized and distributed approaches, but will not specify any algorithm. However, up to now, the resource management support relies on external centralized entities. For example, a centralized path computation engine is used to pre-compute static schedules and paths. Such approaches clearly do not scale up. Our position is to go beyond this limitation by proposing efficient distributed algorithms and implementations for resource allocation.
The objective of this work is to provide the means to operate a WSN with TSCH in a fully distributed manner, namely without using any centralized entities. To reach this goal, the subject focuses on MAC in order to compute and allocate, in a distributed manner, global schedules of cells that will guarantee required properties.