Peak-to-Average Power Ratio Reduction Techniques In OFDM/SC-FDMA Systems

Orthogonal frequency division multiplexing has high spectral efficiency and is robust against multipath fading, and therefore it is incorporated in many practical communication standards, such as the Third Generation Partnership Project Long Term Evolution(3GPP-LTE) and IEEE 802.11a/g wireless standards, etc. However, one of the main drawbacks associated with OFDM is the high peak-to average power ratio(PAPR). To reduce the PAPR of the transmitted signal, the 3GPP-LTE standard has adopted a special form of “linearly precoded” OFDM in its uplink transmission, called single carrier frequency division multiple access(SC-FDMA). The PAPR of SC-FDMA signals is much lower than that of OFDM signals, however, it increases with the increase in the PAPR of the applied constellation. The PAPR of SC-FDMA signals can be very large especially when high order modulations are applied for the future wireless communication systems. The signals with large PAPR will cause significant in-band distortion and undesired out-of-band radiation when it is passed through a transmit power amplifier, which affects the performance of the communication systems greatly and induces interference to other systems adjacent in frequency. To mitigate the nonlinear distortions of the system, it is always necessary to impose input power backoff on the transmitted signal which reduces the power efficiency of the power amplifier greatly. Focusing on the PAPR problem in OFDM/SC-FDMA systems and considering the tradeoff between the PAPR reduction performance, data rate loss, computational complexity and BER performance, the dissertation deeply analyzes and investigates practical and effective PAPR reduction techniques in OFDM/SC-FDMA systems. The main work and contributions of this dissertation are listed as follows:1. For OFDM and SC-FDMA systems, we first give the system models and signal models of them, respectively. Based on the signal models of OFDM and SC-FDMA systems, the definitions of the PAPR are given and the causes for forming large PAPR and the distributions of the PAPR are also analyzed for both systems. Focusing on the PAPR problem in OFDM/SC-FDMA systems, the analysis of performance improvement of the system by performing PAPR reduction is given in terms of the effects of the PAPR reduction on the energy efficiency and spectral efficiency and the saving of the power of the system.2. In the reduction of the PAPR of OFDM systems, we propose the blind SLM method with low computational complexity. By using the frame structure of practical communication systems, one or several OFDM symbols are treated as a processing block, such that in each processing block, there is one OFDM symbol that has pilot subcarriers. The same phase rotation vectors are applied for each processing block to generate candidate signal blocks and then the one with the lowest PAPR is selected for transmission. By delicately designing the phase rotation vectors and replacing the inverse fast Fourier transform operations(IFFT) in generating the candidate signal blocks by the multiplication operations between the conversion matrices corresponding to the phase rotation vectors and the original signal block, the proposed SLM method has very low computational complexity and can enable equivalent channel estimation with the received pilot data to achieve blind data detection without side information at the receiver. The proposed scheme has very low computational complexity and dose not need the transmission of side information, which provides a very promising frame of SLM for practical applications.3. In the reduction of the PAPR of OFDM systems, the semi-blind SLM method with low-complexity transceiver is investigated. For single-input single-output systems, the semi-blind SLM scheme with low-complexity transceiver is proposed where the side information is embedded at the pilot subcarriers. In the proposed scheme, the phase rotation vectors with special forms are designed at the transmitter, which enables generating of the candidate signals with low computational complexity and embedding of the side information(SI) into the pilot subcarriers at the transmitter. At the receiver, pilot-aided embedded SI detection algorithm is introduced to enable data detection without SI. Analysis and simulation results show that the proposed scheme can achieve much lower computational complexity than that of the available semi-blind SLM schemes at both the transmitter and receiver, and almost the same bit error rate(BER) performance as compared with the conventional SLM with perfect SI at the receiver. For the Alamouti coded multiple-input multiple-output(MIMO) OFDM systems with two transmitting antennas, the semi-blind SLM scheme with low-complexity transceiver is proposed where the SI is embedded into the data subcarriers. In the proposed method, the low-complexity candidate signal generating scheme where the SI is embedded into the data subcarriers is investigated, and the low-complexity SI detection algorithm with the received data is also given at the receiver. Analysis and simulation results show that the proposed scheme can achieve better SI detection performance and almost the same PAPR reduction and BER performance with much lower complexity at both the transmitter and receiver, as compared to the available semi-blind SLM schemes.4. In the reduction of the PAPR of SC-FDMA signals, we investigate the simple and flexible PAPR reduction method with low computational complexity and zero excess bandwidth. For the coded SC-FDMA systems, based on the deep analysis of the reason of forming large peaks of SC-FDMA signals, a simple and flexible PAPR reduction method without excess bandwidth is proposed. The proposed method reduces the PAPR by introducing few bit errors to modify the few complex modulated symbols of each data SC-FDMA symbol in a sub-frame, which cause peaks of the output signal samples to be larger than a predetermined threshold value. In addition, the effect on BER performance of the few deliberately corrupted bits can be greatly mitigated by using the channel decoding at the receiver. The proposed scheme has very low computational complexity and can reduce the PAPR of SC-FDMA signals effectively. For the conventional SC-FDMA systems, to avoid the excess bandwidth and noise enhancement penalty involved in the pulse shaping method for PAPR reduction of SC-FDMA signals, we propose a novel alternative method, called probabilistic pulse shaping. In the proposed method, a set of weighting windows without excess bandwidth is designed which can generate candidate signals with very low computational complexity. The SC-FDMA symbols in a slot is treated as a processing block and the candidate signal blocks for each processing block are generated with the designed weighting windows, such that the effect of the weighting windows can be removed in the equalization process after performing the estimation of the equivalent channel at the receiver. It is shown by simulations that the proposed method achieves good performance in terms of both PAPR reduction and BER with low additional complexity at the transmitter.