Banca de DEFESA: JOSÉ MARTINS CASTRO NETO

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : JOSÉ MARTINS CASTRO NETO
DATE: 10/08/2020
TIME: 14:00
LOCAL: meet.google.com/xgs-wgic-cqz
TITLE:

Coexistence Solutions for LTE and Wi-Fi Systems in Multicell Deployments

KEY WORDS:

coexistence, LTE-U, multicellular system, Wi-Fi, machine learning

PAGES: 50
BIG AREA: Engenharias
AREA: Engenharia Elétrica
SUBÁREA: Telecomunicações
SPECIALTY: Sistemas de Telecomunicações
SUMMARY:

The growth of mobile internet access from fourth generation (4G) devices, combined with increasing usage of smartphones, the upcoming fifth generation (5G) and massive usage of multimedia services, make the demand for mobile traffic reach high levels and the need for bandwidth grows. However, the electromagnetic spectrum utilized by these applications is limited, creating scarcity in the face of demand, besides the high costs and bureaucracy for granting new bands. To overcome this problem, one of the solutions is to leverage the unlicensed spectrum, for it is free of charges, presents some of its portions with bandwidth higher than the licensed spectrum, and portions with underused profile, indicating less probability of interference between technologies. In this context rises the LTE-U and LTE-LAA technologies with modifications in the medium access mechanism of LTE for leveraging the unlicensed spectrum in the Industrial, Scientific and Medical (ISM) 5 GHz band. These technologies must coexist with the most successful and consolidated technology already using this portion of the spectrum, the Wi-Fi. However, each technology implements its access mechanism, then studies must be done to point out impacts that each of these technologies has when coexist. Besides the coexistence study, the application of machine learning techniques to automatically adjust the medium access parameters, controlling the generated impact of one technology into another must be realized.   Studies in such a scenario and with single-cell environments have already been explored in literature, remaining the challenge of new solutions targeting multi-cell environments. For all it has been exposed, this work has the following goals: (1) The coexistence study of LTE-U and Wi-Fi technologies in a multi-cell scenario, with co-channel and inter-RAT (same technology) interference; (2) The application of machine learning algorithms (reinforcement learning) to adjust the parameters targeting optimizing the medium access for one, or both technologies, and consequently reach improvements in the coexistence measured in the form of data rates and decreasing packet losses. 

BANKING MEMBERS:
Interno - 2579664 - ALLAN DE MEDEIROS MARTINS
Externo à Instituição - LEONARDO AGUAYO - UnB
Presidente - 1412682 - VICENTE ANGELO DE SOUSA JUNIOR