Thesis
“MAPEO DE QoS ENTRE LAS CAPAS DE ACCESO AL MEDIO Y TRANSPORTE EN SISTEMAS CELULARES 4G”
Autor
LÓPEZ CRUZ, FLORENCIO
Institución
Resumen
ABSTRACT
Nowadays there are multiple services with dynamic content and multimedia real-time
restrictions populating the Internet. These services were initially supported only by terrestrial
networks, however, they are currently available to users of 3G and 4G cellular systems (3rd and
4th generation). 4G also known as Long Term Evolution (LTE) cellular system, pays special
emphasis on the provision of multimedia applications, which claim higher demands for quality of
service (QoS) (i.e, reduced delays, increased spectral efficiency, higher data rates and lower
packet transmission loss).
Multimedia services (voice, video, web, mail, etc.) have differing Quality of Service (QoS)
requirements. QoS is defined as the ability of a network to provide guaranteed resources and
differentiate services. This distinction has led to a performance problem in the provision of
service resources, which LTE tackles by differentiating subscribers. When differentiating by
type or group of subscribers, packet flows assigned to the same bearer are treated with the same
packet transmission characteristics. The packets that are given different treatments of
transmission are sent by separate bearers,recognizable by a QCI (QoS Class Identifier).
Since this evolution of mobile networks is based entirely on IP, and therefore in packet
switching, these demands become QoS requirements of LTE, resolved through inter-operation of
a two service models: Differentiated Services (DiffServ) implanted in the Core Network, and
Class of Service Identifier (QCI) in the radio segment. Both models do not have classes of
service directly comparable. In these conditions it would not be possible to establish a link
between them in the system, because it is not possible to know directly what class of service of
one model to connect to another class of service in the second model.
In this thesis we develop a module that finds correlations between DiffServ and QCI classes. The
module implements an absolute correlation algorithm and a relative correlation algorithm.
Absolute correlation is found if the parameters` values are identical, if they are not, relative
correlation estimation is performed. This last operation is based on concepts of membership
functions of a set of elements A, on a set of elements B. The conditions for estimating the degree
xiii
of similarity of a QCI parameter with one of DiffServ are defined by means of Maximum Delay,
and Maximum Packet Loss parameters.
The module displays a front panel with provisions to select the requested service, estimate
correlations, and a display the result. The module can be expanded with more entries in one or
both quality of service catalogs, as new service requirements emerge, and based on the
correlation concept, find matches for the new services.