Title :
Decentralized Computation of Homology Groups in Networks by Gossip
Author :
Muhammad, Abubakr ; Jadbabaie, Ali
Author_Institution :
Univ. of Pennsylvania, Philadelphia
Abstract :
In this paper, we present an approach towards the computation of certain topological invariants in real sensor networks. As shown by many researchers, these invariants are relevant for modeling certain properties of the network such as coverage and routing. What has been lacking so far is a concrete decentralized method to compute these invariants for proper implementation. In this paper, we give an approach towards such an implementation. The main tools being used here are the the so-called higher order Laplacian operators and distributed methods for their spectral analysis that resemble gossip algorithms.
Keywords :
distributed processing; group theory; spectral analysis; telecommunication network routing; telecommunication network topology; wireless sensor networks; concrete decentralized method; decentralized computation; distributed method; gossip algorithm; higher order Laplacian operator; homology groups; network coverage; network routing; real sensor networks; spectral analysis; topological invariants; Cities and towns; Computer networks; Control systems; Distributed computing; Laplace equations; Matrix decomposition; Network topology; Random processes; Routing; Sensor systems;
Conference_Titel :
American Control Conference, 2007. ACC '07
Conference_Location :
New York, NY
Print_ISBN :
1-4244-0988-8
Electronic_ISBN :
0743-1619
DOI :
10.1109/ACC.2007.4283133
