Distributed Seeking of Nash Equilibria With Applications to Mobile Sensor Networks
2012 (English)In: IEEE Transactions on Automatic Control, ISSN 0018-9286, E-ISSN 1558-2523, Vol. 57, no 4, 904-919 p.Article in journal (Refereed) Published
We consider the problem of distributed convergenceto a Nash equilibrium in a noncooperative game where the playersgenerate their actions based only on online measurements oftheir individual cost functions, corrupted with additive measurementnoise. Exact analytical forms and/or parameters ofthe cost functions, as well as the current actions of the playersmay be unknown. Additionally, the players’ actions are subjectto linear dynamic constraints. We propose an algorithm basedon discrete-time stochastic extremum seeking using sinusoidalperturbations and prove its almost sure convergence to a Nashequilibrium. We show how the proposed algorithm can be appliedto solving coordination problems in mobile sensor networks,where motion dynamics of the players can be modeled as: 1) singleintegrators (velocity-actuated vehicles), 2) double integrators(force-actuated vehicles), and 3) unicycles (a kinematic modelwith nonholonomic constraints). Examples are given in which thecost functions are selected such that the problems of connectivitycontrol, formation control, rendezvous and coverage control aresolved in an adaptive and distributed way. The methodology isillustrated through simulations.
Place, publisher, year, edition, pages
2012. Vol. 57, no 4, 904-919 p.
Convergence, extremum seeking, learning, mobile sensor networks, multi-agent control, Nash equilibrium, noncooperative games, stochastic optimization
Research subject SRA - ICT
IdentifiersURN: urn:nbn:se:kth:diva-72945DOI: 10.1109/TAC.2011.2174678ISI: 000302499500007ScopusID: 2-s2.0-84859722327OAI: oai:DiVA.org:kth-72945DiVA: diva2:488345
FunderKnut and Alice Wallenberg FoundationSwedish Research CouncilICT - The Next Generation
QC 201205082012-02-012012-02-012013-04-11Bibliographically approved