Uma proposta de algoritmo de escalonamento baseado em lógica nebulosa para redes LTE (Long Term Evolution)

Abstract

In recent decades, the increasing use of mobile technologies has motivated the development of new techniques and algorithms to provide high transmission rates in mobile networks. Among them, LTE (Long Term Evolution) technology is one of the most significant standards for fourth generation (4G) mobile telephony. Due to its high data rates, LTE becomes very attractive to several kinds of applications. However, the quality of the transmissions can be severely impacted by the radio resources scheduling process, since it can lead to fluctuations in the delay and in the application data rates. To perform resource allocation in LTE networks, the scheduler uses various parameters estimated from the radio environment. This process can lead to the occurrence of erroneous estimates, which should be mitigated in order to deal with the inaccuracies of the wireless environments. In this context, this master thesis presents a fuzzy-based downlink scheduler for LTE networks, named PAFS (Performance-Aware Fuzzy Scheduler). The results show that the proposed scheduler promotes a suitable allocation of the radio resources, improving the performance of different QoS (Quality of Service) parameters without compromising the fairness among the system users.