Abstract-Some recent communication systems like DVB-T2 standard set up a PAPR reduction technique followed by a linearization's one. So in such a scenario, the performance of the linearization is influenced undoubtedly by the PAPR reduction method. In this paper, we revisit the EVM metric and evaluate a closed form regarding the performance of both the PAPR reduction technique and the linearization's one. We choose the predistortion as a linearization technique and define a predistortion error. Assuming that the baseband OFDM signal is characterized as a complex Gaussian process, we consider the three top categories of PAPR reduction methods presented in [2] and we first study the distribution of the resulted signal. Then, we derive some theoretical expressions of the first and second order moments of the predistortion error and show that the error depends mainly on the PAPR of the signal after PAPR reduction method and on the predistortion quality. Some simulations compared to our proposed model confirm our results.

No abstract

International audienceIn this study, the authors look into digital predistortion and its combination with the peak-to-average power ratio (PAPR) reduction technique in the case of orthogonal frequency division multiplexing (OFDM)-based systems. First of all, the authors propose a new predistortion methodology which uses the addition of signals. Two different algorithms are investigated for the predistortion signals generation. The first one is based on the traditional predistortion method, where an estimation of the power amplifiers (PAs) characteristics is needed in the first place. The second algorithm to generate the predistortion signal does not need a priori PA estimation. It is based on an iterative compensation of the error between the amplified signal and the OFDM signal. Then, the authors combine this latter algorithm with the tone reservation PAPR reduction method in order to have a global adding signal expression. Two combination scenarios of PAPR reduction and predistortion obtained by means of adding a signal are proposed. The first one combines both of them in series while the second one is parallel. Performances of the proposed predistortion algorithms and the combination scenarios are compared thanks to simulations based on IEEE802.il a/g standards for a memoryless solid state PA

No abstract

No abstract