Análise de desempenho de sistemas multiportadora em canais com ceifamento de pico

Abstract

The advent of Digital Television has brought a new panorama for digital telecommunications, presenting new problems and challenges to be addressed effectively. Today, there are three different standards in operation: the ATSC developed in the USA, the DVB-T developed in Europe and the ISDB-T developed in Japan. Among these standards, DVB-T and ISDB-T use a transmission technique with (ORDM) (ortho-gonal Frequency Division Multiplexing). With the adoption of this technique in European and Japanese digital television standards, there has been a significant increase in interest in the implementation of OFDM. Among the major difficulties encountered, two are worth mentioning: the complexity of constructing several oscillators in phase and the problem of signal harvesting by the power amplifier. The first problem occurs because this type of transmission uses a large number of carriers that need to be in phase to guarantee the orthogonality between them. With the development of digital signal processors, it was possible to solve this problem by using IFFT for modulation and FFT for demodulation. The second problem occurs because the OFDM signal has a high relation between the peak power and the average power, that is, this signal can have high amplitude peaks. As power amplifiers are designed to operate near the saturation point, in order to provide high performance, these peaks are harvested, introducing a non-linear distortion in the signal. The objective of this work is to first present the methods used for generation and reception of OFDM signals and to study their characteristics, determining the reasons for which the harvesting occurs. Based on these characteristics, two methods for modeling the harvesting are presented, obtaining the probability of error in both cases. Also a comparison is made between the obtained results, defining the most adequate model to estimate the probability of symbol error due to the harvest.