Emergence of Techno-Social Norms in Cognitive Radio Environments

TL;DR

This paper investigates how game theory can be used to identify and embed techno-social norms in cognitive radio systems, potentially regulating spectrum use without external enforcement.

Contribution

It introduces the concept of techno-social norms derived from game theory simulations to regulate unlicensed spectrum access in cognitive radio environments.

Findings

Emergence of equilibria mirroring behavioral trends

Norms can be embedded into cognitive radios for regulation

Simulations provide insights into autonomy and regulation balance

Abstract

The aim of this paper is to explore the potential of Game Theory (GT) in extracting rules of behaviour for emerging Cognitive Radio environments. We revisit the commons approach to unlicensed spectrum and try to show that a commons can be basically regulated from the inside out. GT simulations of CR interactions reveal the emergence of certain equilibria mirroring behaviours/trends?. Once these ?trends identified, norms may be expressed and then embedded into machines (CRs). Internalized norms may thus become the alternative to external enforcement of rules. We call these emerging norms techno-social norms (TSNs). TSNs could eventually become a means of regulating the use of unlicensed spectrum and making open spectrum access feasible. Open spectrum access scenarios are considered and analysis is performed based on reformulations of two game theoretical models: Cournot and Bertrand. The standard oligopoly models are reformulated in terms of radio resource access in unlicensed bands. In order to capture the large variety of CR interaction situations, several GT equilibrium concepts are considered: Nash, Pareto, Berge-Zhukovskii, and Lorenz. In order to capture the heterogeneity of CR interactions, the standard GT model is enriched allowing players to be biased toward different types of equilibrium (or rationality). An evolutionary game-equilibrium detection method is used. Numerical simulations bring relevant insights on the problem of autonomy vs. regulation in emerging CR environments. Relying on extensive GT simulations, some rules of behaviour - to be expanded into techno-social norms - may be derived.