Fast Convergence of Natural Bargaining Dynamics in Exchange Networks

Authors:

Abstract
Bargaining networks model the behavior of a set of players that need to reach pairwise agreements for making profits. Nash bargaining solutions are special outcomes of such games that are both stable and balanced. Kleinberg and Tardos proved an algorithmic characterization of such outcomes, but left open the problem of how the actual bargaining process converges to them. A partial answer was provided by Azar et al. who proposed a distributed algorithm for constructing Nash bargaining solutions, but without polynomial bounds on its convergence rate. In this paper, we introduce a simple and natural model for this process, and show fast convergence to approximate Nash bargaining solutions. Consider an exchange network represented by a weighted graph. At each time step, each player proposes a deal to each of her neighbors. The proposal consists of a share of the potential profit in case of agreement. The share is chosen to be balanced in Nash's sense as far as this is feasible (with respect to the current best alternatives for both players). We show that fixed points of our dynamics correspond to Nash bargaining solutions when such solutions exist, and that \eps-approximate fixed points correspond to approximate Nash bargaining solutions for small enough \eps. Crucially, we also prove that our dynamics converges to an \eps-approximate fixed point in time O(\eps^{-2}) independent of the network size when the potential earnings (weights) are uniformly bounded. Moreover, for small enough \eps, the best pairing between the nodes emerges. As an example, we prove that for bipartite graphs this leads to polynomial time convergence to an approximate Nash bargaining solution, with probability close to one under small random perturbations.

This publication has 10 references indexed in Scilit: