Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Membrane Computing Approach to the Generalized Nash Equilibrium

Alejandro Luque-Cerpa, Miguel A. Gutiérrez-Naranjo

TL;DR

This work tackles computing Generalized Nash Equilibria ($GNE$) in population games by bridging Evolutionary Game Theory ($EGT$) with Membrane Computing. It proposes a first approach that models $GNEP$ via Transition P systems with membrane polarization and embeds Brown‑von Neumann‑Nash ($BNN$) dynamics into a discretized, five‑stage EDM‑PDM framework, analyzed for linear time complexity in the time horizon. The paper provides a formal P‑system design and a complexity proof, complemented by a small experimental demonstration in an Energy Market Game showing convergence to a $GNE$ within a few iterations. The significance lies in a novel computational bridge between $GNEP$ and Membrane Computing, with potential for applying membrane‑based methods to a broad class of EDM‑PDM problems and real‑world equilibrium computations.

Abstract

In Evolutionary Game Theory (EGT), a population reaches a Nash equilibrium when none of the agents can improve its objective by solely changing its strategy on its own. Roughly speaking, this equilibrium is a protection against betrayal. Generalized Nash Equilibrium (GNE) is a more complex version of this idea with important implications in real-life problems in economics, wireless communication, the electricity market, or engineering among other areas. In this paper, we propose a first approach to GNE with Membrane Computing techniques and show how GNE problems can be modeled with P systems, bridging both areas and opening a door for a flow of problems and solutions in both directions.

A Membrane Computing Approach to the Generalized Nash Equilibrium

TL;DR

This work tackles computing Generalized Nash Equilibria () in population games by bridging Evolutionary Game Theory () with Membrane Computing. It proposes a first approach that models via Transition P systems with membrane polarization and embeds Brown‑von Neumann‑Nash () dynamics into a discretized, five‑stage EDM‑PDM framework, analyzed for linear time complexity in the time horizon. The paper provides a formal P‑system design and a complexity proof, complemented by a small experimental demonstration in an Energy Market Game showing convergence to a within a few iterations. The significance lies in a novel computational bridge between and Membrane Computing, with potential for applying membrane‑based methods to a broad class of EDM‑PDM problems and real‑world equilibrium computations.

Abstract

In Evolutionary Game Theory (EGT), a population reaches a Nash equilibrium when none of the agents can improve its objective by solely changing its strategy on its own. Roughly speaking, this equilibrium is a protection against betrayal. Generalized Nash Equilibrium (GNE) is a more complex version of this idea with important implications in real-life problems in economics, wireless communication, the electricity market, or engineering among other areas. In this paper, we propose a first approach to GNE with Membrane Computing techniques and show how GNE problems can be modeled with P systems, bridging both areas and opening a door for a flow of problems and solutions in both directions.
Paper Structure (10 sections, 25 equations, 1 figure, 1 algorithm)

This paper contains 10 sections, 25 equations, 1 figure, 1 algorithm.

Table of Contents

  1. Introduction
  2. Population Games under BNN Dynamics
  3. Transition P systems with membrane polarization
  4. Design and functioning of the P System
  5. Overview of the computation
  6. Experiment
  7. Conclusions
  8. Russian peasant multiplication in Membrane Computing
  9. Computation Analysis
  10. Definition of the P system

Figures (1)

  • Figure 1: Small experiment with three players and five strategies, distributed such that $S^1 = \{3,5\}$, $S^2 = \{1,3,5\}$, $S^3 = (S^1)^c = \{1,2,4\}$. Players 1, 2, and 3 are represented by red, blue, and green lines respectively. After five time steps, the GNE is reached. It was expected that player 3 would concentrate more agents over time slot 4 ($z_4^3$) or 2 ($z_2^3$) since it is the only player with access to those resources. The difference is a consequence of the cost difference. The key observation in this plot is that $\dot{z} = 0$ after five time steps.