Controle de distorções em sistemas OFDM com amplificadores de potência não lineares

Abstract

The OFDM technology has emerged as an important tool for dealing with the frequency selectivity of mobile channels that causes intersymbol interference. Due to this fact and to many other interesting properties, OFDM was chosen for the 5G (5th Generation), as it happened with the 4G (4th Generation). One of the main disadvantages of OFDM systems is the high sensitivity to nonlinear power amplifiers. When the OFDM signal amplitude peaks reach the PA saturation region, the signal is affected by distortions that compromise the transmission quality. Another important technology in the wireless communications scenario is the cooperative networks. This arrangement of network devices ensures the cooperation between users within the cell, promoting an increase in the coverage area and, consequently, an improvement in transmission quality. In addition to cooperative networks and OFDM systems, radio resource management in OFDM systems has proven to be an important tool to improve the performance of mobile communication systems. Resource allocation considering the presence of nonlinear distortions inserted by PAs is not yet a well-developed research area in the literature, however, it can provide considerable network performance and significantly reduce the amount of distortions present in the OFDM signal. In this work, we present two approaches for the treatment of nonlinear distortions in OFDM systems. In the first approach, we propose a set of techniques for reducing peaks in a cooperative OFDM system with relay AF and non-linear PAs. The set of techniques is based on the PTS technique and PAPR and CM metrics are used as to select the signals transmitted. The proposed techniques aim to reduce the peaks of the signal transmitted by both the source and the relay simultaneously. In the second approach, we propose a power allocation problem in a non-cooperative OFDM system taking into account the distortions inserted by a third-order polynomial PA. Based on this formulation, two sub-optimal heuristics are proposed for power allocation in the considered scenario. The two heuristics are based on the optimal power allocation algorithm for the linear case: Hughes-Hartogs (HH). We also propose complexity relaxation solutions for the sub-optimal algorithms and, finally, we perform an analysis of the algorithmic complexity of the methods. Simulation results are presented to evaluate the performance of the set of proposed techniques.