A Membrane Computing Approach to the Generalized Nash Equilibrium

Alejandro Luque-Cerpa,Miguel A. Gutiérrez-Naranjo

2024-04-27

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.

Computer Science and Game Theory

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: How to use Membrane Computing technology to solve the Generalized Nash Equilibrium (GNE) problem. Specifically, the author proposes a membrane - computing method based on P - systems to simulate and solve the Generalized Nash Equilibrium problem under Brown - von Neumann - Nash (BNN) dynamics.

### Problem Background

1. **Nash Equilibrium and Generalized Nash Equilibrium**
- In Evolutionary Game Theory (EGT), when no individual in a group can increase its payoff by changing its strategy alone, the group reaches Nash Equilibrium. This means that in this state, no individual can get a better result by defecting.
- Generalized Nash Equilibrium (GNE) is a more complex version of Nash Equilibrium, in which each player's strategy set may depend on the strategies of other players. GNE has important applications in economics, wireless communication, power markets, and engineering, etc.

2. **Membrane Computing and P - systems**
- Membrane Computing is a computational model inspired by intracellular biochemical reactions, and P - systems are the main tools in membrane computing. P - systems have been successfully applied to model many real - life dynamic processes.
- This paper proposes a new P - system - transition P systems with membrane polarization - to solve the Generalized Nash Equilibrium problem.

### Main Contributions of the Paper

1. **Proposing a New Method**
- The author applies membrane - computing technology to the Generalized Nash Equilibrium problem for the first time and shows how to use P - systems to model GNE problems.

2. **Linear Complexity**
- It is proved that the proposed P - system has linear complexity in time and does not depend on the number of players or the number of strategies.

3. **Experimental Verification**
- The effectiveness of this method is verified through a small - scale experiment, showing that the P - system can quickly converge to the Generalized Nash Equilibrium.

### Formula Representation

The formulas involved in the paper include:

- BNN dynamic equations:
\[
\dot{x}_i^k(t)=m_k [\hat{p}_i^k(t)]^+ - x_i^k(t)\sum_{j\in S_k} [\hat{p}_j^k(t)]^+
\]
\[
\hat{p}_j^k(t)=p_j^k(t)-\frac{1}{m_k}\sum_{l\in S_k}x_l^k(t)\cdot p_l^k(t)
\]

- Payoff function update:
\[
p(t)=f(x(t)) = f(M\cdot z(t))=-S\cdot M\cdot z(t)-C^\top\bar{J}-\alpha\odot(M\cdot z(t))-\beta
\]

These formulas are used to describe the strategy evolution and payoff update processes in group games.

### Summary

This paper aims to provide a novel method to solve the Generalized Nash Equilibrium problem through membrane - computing technology, filling the gap between Evolutionary Game Theory and membrane computing and providing a new direction for future research.