Page 66 - Proceedings of the 2018 ITU Kaleidoscope

MULTIFRACTAL MODELING OF THE RADIO ELECTRIC SPECTRUM APPLIED IN
COGNITIVE RADIO NETWORKS

1
Luis Tuberquia-David ; Cesar Hernandez 1
1
Universidad Distrital Francisco Jose de Caldas, Colombia

ABSTRACT Communication networks have been a subject of study for
over two decades, showing that the times between the arrival
The work discussed in this article is framed within the of user demands and the demand placed on the network have
context of cognitive networks in America, showcasing the a correlation that persists across different time scales [8].
scenario of the radioelectric spectrum of the city of Bogotá, Since the 1980's, studies have been developed on data traffic,
Colombia. The objective is to model the traffic of the wireless in order to predict, control and improve the overall service.
network, since it is underused in this region of Latin America. Furthermore, the research carried out at Bellcore [9] on
Hence, some tools are studied to allow the structuring of the Ethernet traffic revealed that the nature of Ethernet traffic is
type of traffic seen in the network. Based on stochastic tools multifractal [10]. Additionally, it has been pointed out that
such as the log-scale diagram, the linear multiscale diagram, the time series of incoming packets related to user demand
and the multifractal spectrum, this research aims to verify behave with stochastic self-similarity which is the main
the multifractality of traffic series collected on the electric characteristic of multifractal signals [11], [12]. Moreover,
radio spectrum of Bogotá, Colombia in 2012. In fact, the multifractal traffic has been validated in different network
study reveals that all the channels of the network have a structures as evidenced in [13], [14].
multifractal behavior with 90% of them presenting a Hurst
parameter in the 0.5 to 1 range. The evidence suggests that The use and demand of wireless data have increased recently,
the traffic in this region could be modeled as multifractal which is expected to continue due to the nonstop
time series. Therefore, the analysis carried out intends to development of wireless applications. This means that the
provide a new modeling method for the Colombian monthly use of mobile data will be multiplied by eight before
radioelectric spectrum in the form of a multifractal-based 2020 in comparison to 2015 [15]. The research detailed in
analysis. this article is based on the three mentioned premises: the
underuse of the radioelectric spectrum in Colombia, the
Keywords – Multifractal, cognitive radio, radio electric multifractal nature of network traffic and the growing trend
spectrum, Bogotá, Hurst Parameter of wireless systems. Therefore, the main goal of the authors
is to confirm the nature of the traffic data captured in the city
1. INTRODUCTION of Bogotá within the requested Wi-Fi band. Furthermore,
probabilistic tools are used such as the Log-scale Diagram
Historically, spectrum allocation in Colombia is regulated by (LD) to calculate the Hurst parameter, The Multiscale-linear
the Ministry of Information and Communications Diagram (MD) to estimate said parameter in two different
Technologies (by its acronym in Spanish, TIC), with the help statistical moments, and the Multifractal Spectrum (MS)
of the National Agency of the Spectrum (by its acronym in analysis to determine the width of the spectrum. Hence, the
Spanish, ANE). This branch is in charge of granting users the research wishes to prove that the network traffic in Bogotá
best quality in service, although studies such as [1], [2] have has a multifractal behavior.
revealed that the radio spectrum in Colombia is underused.
Many research projects have been carried out around the To reach the proposed objective, spectral occupancy data
globe, stating that the radioelectric spectrum is underused in corresponding to the Wi-Fi band was used since it is freely
terms of frequency, time and geographic domains [3],[4]. available and has a more chaotic behavior which makes it
The results of the investigations show that this is a global harder to model compared to the GSM band.
phenomenon that not only occurs in Colombia. As a solution
to the problem of underuse, several solutions have been In Section 2, the article starts by explaining the data
proposed from where cognitive radio (CR) networks [5] collection process and how the time series was built. In
stand out. CR presents a dynamic spectrum management Section 3, the steps required to compute the MS are detailed.
technique, known as Dynamic Spectrum Access (DSA) [6] Section 4 presents the results of the statistical analysis of the
which is designed to prevent interference, adapt to the data gathered from the radioelectric spectrum of Bogotá. The
immediate availability of the local spectrum and create times conclusions of this investigation are stated in Section 5.
and locations for Secondary Users (SU) to share with
Primary Users (PU) [7].

978-92-61-26921-0/CFP1868P-ART @ 2018 ITU – 50 – Kaleidoscope