A payoff-based learning procedure and its application to traffic games

Cominetti, Roberto; Melo, Emerson; Sorin, Sylvain

doi:10.1016/j.geb.2008.11.012

Cited by 106 publications

(139 citation statements)

References 35 publications

Supporting

Mentioning

130

Contrasting

Order By: Relevance

“…Here we extend the learning algorithm to the generalized spatial congestion games on any generic interference graphs with heterogeneous users, which lead to significant differences in analysis. For example, we show that the convergence condition for the learning algorithm depends on the structure of spatial reuse, which is different from those results in [28], [29].…”

Section: B Distributed Learning Algorithmcontrasting

confidence: 90%

“…The idea is to extend the principle of single-agent reinforcement learning to a multi-agent setting. Such multi-agent reinforcement learning algorithm has also been applied to the classical congestion games on complete interference graphs [28], [29] by assuming that users are homogeneous (i.e., user's payoff only depends on the number of users choosing the same resource). Here we extend the learning algorithm to the generalized spatial congestion games on any generic interference graphs with heterogeneous users, which lead to significant differences in analysis.…”

Section: B Distributed Learning Algorithmmentioning

confidence: 99%

“…Case 3: If player N +1 can generate interference to a player n ∈ N in the original game, then player N + 1 can compute its best response strategy a * N +1 according to (29) if player n can also generate interference to player N + 1, and according to (28) otherwise. By checking that whether a * N +1 and a * n are the same, we have the following two possibilities.…”

Section: B Proof Of Lemmamentioning

confidence: 99%

“…This proof is derived based on the one in [29]. The key difference is that we consider that the generalized spatial congestion games on any generic graphs and users can be heterogenous.…”

Section: Poa≥mentioning

confidence: 99%

“…The key difference is that we consider that the generalized spatial congestion games on any generic graphs and users can be heterogenous. While, [29] only considers the special case that the congestion game is defined over the complete graphs and users are homogenous.…”

Section: Poa≥mentioning

confidence: 99%

See 4 more Smart Citations

Spatial Spectrum Access Game

Chen

Huang

2017

Handbook of Cognitive Radio

View full text Add to dashboard Cite

Abstract-A key feature of wireless communications is the spatial reuse. However, the spatial aspect is not yet well understood for the purpose of designing efficient spectrum sharing mechanisms. In this paper, we propose a framework of spatial spectrum access games on directed interference graphs, which can model quite general interference relationship with spatial reuse in wireless networks. We show that a pure Nash equilibrium exists for the two classes of games: (1) any spatial spectrum access games on directed acyclic graphs, and (2) any games satisfying the congestion property on directed trees and directed forests. Under mild technical conditions, the spatial spectrum access games with random backoff and Aloha channel contention mechanisms on undirected graphs also have a pure Nash equilibrium. We also quantify the price of anarchy of the spatial spectrum access game. We then propose a distributed learning algorithm, which only utilizes users' local observations to adaptively adjust the spectrum access strategies. We show that the distributed learning algorithm can converge to an approximate mixed-strategy Nash equilibrium for any spatial spectrum access games. Numerical results demonstrate that the distributed learning algorithm achieves up to 100% performance improvement over a random access algorithm.

show abstract

Section: B Distributed Learning Algorithmcontrasting

confidence: 90%

Section: B Distributed Learning Algorithmmentioning

confidence: 99%

Section: B Proof Of Lemmamentioning

confidence: 99%

“…This proof is derived based on the one in [29]. The key difference is that we consider that the generalized spatial congestion games on any generic graphs and users can be heterogenous.…”

Section: Poa≥mentioning

confidence: 99%

Section: Poa≥mentioning

confidence: 99%

See 3 more Smart Citations