Ioanis Nikolaidis
Simon Fraser University
Centre for Scientific Computing Seminar
(in conjunction with PIMS/MITACS)
School of Engineering Science
IEEE Circuits and Systems joint Chapter of the Vancouver/Victoria Sections

Computing Science Department
University of Alberta


(Presentation is available in html, ps, and pdf formats.)

Thursday, August 22, 2002, 1:00 p.m. in East Academic Annex Room 1100


Mobile Ad-Hoc Networks (MANETs) have been proposed as one of the possible solutions to the scalability problem of wireless networks. The ad-hoc nature of MANETs where all nodes have equal rights and responsibilities, suggests the need for fairly complicated coordinated control schemes in order for essential tasks, such as routing, to be accomplished. The node mobility, experienced as the continuous change of the underlying network topology, necessitates that such control schemes be agile and adapt to the new topology all of the time. A number of constraints complete the picture, ranging from interference (limits to the spatial reuse of the frequency spectrum), to congestion (via suitable medium access protocol), energy constraints, and others.

The talk will be geared towards the algorithmic aspects of providing energy-optimal distribution of data in MANETs and broadcast services in particular. Our free variable is the per-node transmission power (expressed, equivalently, as the per-node transmission radius or per-node transmission coverage). The transmission radii are directly related to the energy cost, and the minimization of the global energy cost is our objective. The underlying problem formulated for static (non-mobile) instances, has recently been proven to be a computationally hard problem and several heuristics have been proposed. We will present and compare some of these heuristics. Contrary to previous studies, that deal with static wireless environments, we will then add the mobility dimension and subsequently refine the schemes such that localized actions (without the need for the expensive global heuristics to be re-run) preserve the network connectivity without an undue energy penalty. We find out that unless clever "repair" actions are performed, the near-optimality of the previously proposed heuristics results in extremely fragile connectivity.

[As part of the talk, we will also introduce caveats related to the simulation of MANETs using the Random Waypoint (RWP) mobility model.]


Ioanis Nikolaidis is an Assistant Professor with the Computing Science Department at the University of Alberta. He received his B.Sc. from the University of Patras, Greece, in 1989 and his M.Sc. and Ph.D. from the Computer Science Department at Georgia Tech in 1991 and 1994, respectively. During 1995-96 he worked for the European Computer-Industry Research Center in Munich, Germany. His research interests include the performance evaluation of computer network protocols.

Last updated Wednesday August 21 19:41:10 PDT 2002.