Séminaire (organisé par l’équipe de recherche RO)


Doctorant Heudiasyc

AND - Adaptive Neighbor Discovering in Vanets

Mardi 29 septembre 2015 à 11 h en salle GI041

Résumé :

Networks composed by vehicles (Vanets) constitute very dynamic networks. Vehicles are moving all the time, leading to uninterrupted changes in the network topology. Exchange data through these networks requires that any vehicle first recognize, or discover, other vehicles within its communication range. This discovering process, called Neighbor Discovery (ND), is an imperative initial step to allow many protocols as routing, medium access and topology knowledge work efficiently and correctly. A first solution to ND is make a node send beacon messages as fast as it can, speeding-up discovering process. Unfortunately, this behavior leads to a fast draining of resources - batteries of sensors or smart-phones ; communication spectrum in Vanets ; etc. In fact, perform an efficient neighbor discovering in dynamic networks without spending a lot of resources is still an open task. AND - Adaptive Neighbor Discovering - is a new algorithm developed with the aim of improve neighbor discovering while saving communication spectrum in Vanets. To achieve its goals, AND uses an adaptive inter-beacon delay calculated considering the speed of nodes and their knowledge about their neighborhood. Initial results obtained through simulations were promising, showing improvements in relation to other known algorithms.


FR SHIC 3272

Collegium UTC/CNRS