Domen ŠOBERL: Decentralized computation of homology in wireless sensor networks using spanning trees

☰ Meni

V ponedeljek, 17. februarja 2020, bo ob 16.00  v prostorih Fakultete za matematiko, naravoslovje in informacijske tehnologije Univerze na Primorskem (Glagoljaška 8, Koper) predavanje v okviru PONEDELJKOVEGA SEMINARJA RAČUNALNIŠTVA IN INFORMATIKE Oddelkov za Informacijske znanosti in tehnologije UP FAMNIT in UP IAM.

ČAS/PROSTOR: 17. februar 2020 ob 16.00 v FAMNIT-VP2

———————————————–

PREDAVATELJ: Domen ŠOBERL

Domen Šoberl is a teaching assistant at UP FAMNIT and is currently finishing his PhD thesis at the Faculty of Computer and Information Science in Ljubljana. Research interests: artificial intelligence, agent learning and planning, computational topology.

————————————————————————————————————-

NASLOV: Decentralized computation of homology in wireless sensor networks using spanning trees

POVZETEK:

Wireless Sensor Networks (WSN) are computer networks of small inexpensive sensor nodes with a limited computational power and often without positioning capabilities. They are deployed over an area of interest to perform some measuring tasks and communicate the results. When deployed arbitrarily, the information on signal coverage is crucial to asses the quality of deployment. A homological criteria has been proposed to guarantee a full coverage using only the connectivity data.
In this seminar I will present our decentralized algorithm to verify homology-based WSN coverage criteria. Individual nodes examine the homology of their immediate neighborhood. They organize themselves in a spanning tree formation and merge their local information up the tree. During the process, a simplicial complex known as Vietoris-Rips complex is constructed, which encodes certain information on the shape of the WSN coverage. The algorithm works in arbitrary network topology and its communication demands are suitable for low-bandwidth wireless sensor networks.

Predavanje bo potekalo v angleškem jeziku.

Vabljeni!