Interference and Deployment Issues for Cognitive Radio Systems in Shadowing Environments

TL;DR

This paper models the aggregate interference caused by cognitive radios in shadowing environments, highlighting the limitations of lognormal models and providing deployment guidelines based on interference statistics.

Contribution

It introduces a new interference model considering distance attenuation and fading, and evaluates deployment parameters with and without environment map knowledge.

Findings

Aggregate interference cannot be accurately modeled by a lognormal distribution.

Deployment parameters like exclusion zone radius are derived for different fading environments.

Knowledge of radio environment maps significantly increases feasible cognitive radio deployment.

Abstract

In this paper we describe a model for calculating the aggregate interference encountered by primary receivers in the presence of randomly placed cognitive radios (CRs). We show that incorporating the impact of distance attenuation and lognormal fading on each constituent interferer in the aggregate, leads to a composite interference that cannot be satisfactorily modeled by a lognormal. Using the interference statistics we determine a number of key parameters needed for the deployment of CRs. Examples of these are the exclusion zone radius, needed to protect the primary receiver under different types of fading environments and acceptable interference levels, and the numbers of CRs that can be deployed. We further show that if the CRs have apriori knowledge of the radio environment map (REM), then a much larger number of CRs can be deployed especially in a high density environment. Given REM information, we also look at the CR numbers achieved by two different types of techniques to process the scheduling information.