On Peak Reduction Techniques For OFDM Systems

As one of the leading candidates for future wireless communication systems, orthogonal frequency division multiplexing (OFDM) suffers from the drawback of high peak-to-average power ratio (PAPR), which would reduce the efficiency of the high power amplifier (HPA) and cause severe in-band nonlinear distortion as well as out-of-band distortion. To tackle PAPR reduction problems in OFDM systems, this dissertation focuses on optimizing total system performance from several perspectives, by seeking a tradeoff among factors like PAPR, transmission performance, spectrum efficiency and computational complexity.The first chapter of this dissertation introduces the features of OFDM systems, reviews the state-of-the-art PAPR reduction techniques, summarizes the problems needed to be addressed, and outlines the main contents of this dissertation.In the second chapter, this dissertation analyzes the PAPR of OFDM signals, and clarifies the definition and distribution of PAPR models and their relations with HPA, serving as the theoretical base of the latter chapters.Clipping methods are a class of widely used peak reduction algorithms. In order to reduce peak values and out-of-band distortion simultaneously, transmitted signals are usually processed by iterative clipping-filtering algorithms. The high computational complexities of the conventional iterative-clipping-filtering algorithms make them unsatisfactory for uplink transmission. While peak cancellation methods have low computational complexities, they would bring problems such as in-band distortion and large peak regrowth. The third part of this dissertation proposes an improved peak cancellation method, which could obtain a better tradeoff among computational complexity, peak reduction and in-band distortion. This method is suitable for uplink transmission for its considerably low computational complexity. Furthermore, a receiver structure with distortion reconstruct scheme is introduced to further improve system performance. The nonlinear distortion introduced by the pre-distortion method would interfer pilots and deteriorate channel estimations. Accordingly, this dissertation presents an improved cancellation algorithm with low complexity, which could effectively eliminate nonlinear distortions on pilots and improve the transmission performance in fading channels.Multiple signal representation (MSR), such as partial transmit sequence (PTS), could reduce PAPR without introducing distortion. Therefore MSR is suitable for downlink transmission. However its high computational complexity is the main obstacle of its utilization. The fourth part of the dissertation proposes a modified PTS algorithm, which could efficiently reduce the computational complexity by estimating the maximal envelope of transmitted signals using geometric interpolation. Another drawback of PTS is the requirement of extra side information. To deal with this problem, a blind PTS algorithm is introduced for pilot-aided OFDM systems.The last chapter of the thesis investigates PAPR reductions of multiple access combined OFDM systems. For Orthogonal Frequency Division Multiplexing Code Division Multiple Access (OFDM-CDMA) systems, an enhanced blind PTS algorithm is proposed to increase spectrum efficiency and transmission performance. A distortionless peak reduction algorithm with low computational complexity is also introduced. For localized frequency division multiple access (LFDMA) systems, a selection among available peak reduction methods has been made and the advantages of the peak cancellation algorithm proposed by this dissertation have been verified.