Peak-to-Average Power Ratio Reduction In OFDM Systems

Orthogonal frequency division multiplexing (OFDM) has drawn explosive attention in a number of current and future standard systems including the IEEE802.11a/g wire-less standard and3GPP-LTE, owing to the advantages of high spectral efficiency and ro-bustness to frequency selective fading. Nevertheless, OFDM suffers from a high peak-to-average power ratio (PAPR). A large PAPR requires a linear high power amplifier (HPA), which, however, is inefficiently used. This thesis focuses mainly on the PAPR reduction of OFDM signals, which is organized as follows:1. For distortion PAPR reduction techniques, a novel scheme that transform the o-riginal OFDM signals into piecewise-distributed signals is proposed. In addition, analysis shows that under certain conditions, the proposed scheme without de-companding at the receiver can also offer a good BER performance. Besides, take power limitation into account, a constant average power adaptive companding scheme is also introduced. Then, through the Bussgang theorem and the distortion noise analysis, a trapezoidal-based companding scheme is presented. By properly choosing the parameters, it can achieve an effective tradeoff between PAPR reduc-tion and bit error rate (BER) performance, which gives OFDM system more design flexibility. Finally, by introducing a weighted coefficients with adaptive windowing length, a novel peak windowing method is proposed in order to avoid the adjacent windowing function overlap.2. A novel immune and evolutionary partial transmit sequences (PTS) scheme is in-troduced to reduce the complexity. Moreover, quantum gate is introduced to modify the fitness of the solution, so as to updating the qubit individuals. During the actu-al operation, the immune algorithm refrains degeneracy phenomena arising from the evolutionary process, thus making the fitness of population increase steadi-ly. Then, by utilizing the correlation and Hamming distance among the candidate signals generated in PTS, a new scheme is proposed to decrease the computa-tional complexity. Finally, an novel algorithm based on peak windowing residual noise, which is generated by windowing the signal samples that exceed the prede-termined threshold, is proposed. Filtering the residual noise to satisfy tone reser-vation constraints and scaling with a scaling factor determined to minimize the out-of-range power generates the peak-canceling signal. The proposed scheme achieves significant PAPR reduction while maintaining low complexity.3. In tone injection (TI), given that only constellations located on the outer ring can be shifted, a novel parallel tabu search algorithm is introduced for TI. In addition, a new parameter for further reduce the power increase is also put forth. Then, a new TI scheme that uses the clipping noise to find the optimal equivalent constellations is proposed. By minimizing the mean rounding error of the clipping noise and possible equivalent constellations, the proposed scheme can easily determine the size and position of the optimal equivalent constellations. Simulation results are provided to demonstrate the effectiveness of the proposed algorithm.