Tóm tắt luận văn tốt nghiệp

TRÖÔØNG ÑAÏI HOÏC KYÕ THUAÄT TP HCM

KHOA ÑIEÄN – ÑIEÄN TÖÛ

BOÄ MOÂN VIEÃN THOÂNG



TOÙM TAÉT

LUAÄN VAÊN TOÁT NGHIEÄP

ÖÙNG DUÏNG BOÄ CAÂN BAÈNG

DUØNG NEURAL NETWORKS

TRIEÄT NHIEÃU GIAO THOA KYÙ TÖÏ

TRONG HEÄ THOÁNG GSM

INTER-SYMBOL INTERFERENCE CANCELLATION

FOR GSM SYSTEM USING NEURAL NETWORKS EQUALIZER

GVHD :ThS HOAØNG ÑÌNH CHIEÁN

SVTH :TRÖÔNG AÙNH THU 49600887

LEÂ THANH NHAÄT 49601145

ÖÙng duïng boä caân baèng duøng Neural Networks trieät nhieãu giao thoa kyù töï trong heä thoáng GSM

Khoùa 1996-2001

ABSTRACT

Neural Networks have seen an explosion of interest in the last few years and have been

successfully applied to a wide range of problem domains, in areas as diverse as finance,

medicine, engineering, geology and physics. In fact, wherever prediction, classification,

control and processing are needed, Neural Networks can be introduced. This sweeping

success can be attributed to Neural Networks as the latter have sophisticated techniques

that are capable of modeling extremely complex functions. Neural Networks can learn

from examples. The Neural Networks user only has to gather representative data, then

evokes training algorithms so that the networks can automatically learn the structure of

the data. Training Neural Networks is in fact an adjustment of some parameters of the

networks to minimize square error. Weight and bias are two of the parameters that are

adjusted in the analysis.

The aim of this thesis is to apply Neural Networks to the equalizer for intersymbol

interference suppression in GSM system. To achieve this aim, the thesis deals with two

main parts : (1) Introduction to mobile communications system (GSM), equalizer and

Neural Networks, and (2) Simulating Neural Networks on the computer.

The first part of the thesis gives a general introduction to Neural Networks and some of

their important models. Some fundamental knowledge of mobile communications system

(GSM) and equalizer is also provided in this part.

The second part of the thesis is concerned with computer simulation programmed in

Matlab 5.3. This part offers some model communications in Gauss Noise, Fading,

Cochannel environment and noise suppression techniques by the use of Neural Networks.

Generally speaking, Neural Networks can be used for cancellation of all kinds of noise,

however, because of the time limit of our thesis, we can only apply Neural Networks to

two types of modulation, BPSK and QPSK. Simulation results are presented plotting BER

curves.

Finally, some conclusions are drawn from simulation results and followed by the

suggestions for further developments of the thesis.

Leâ Thanh Nhaät-Tröông AÙnh Thu 2 GVHD : Hoaøng Ñình Chieán