Research Of PAPR Reduction In OFDM System Based On PTS Algorithm

Orthogonal frequency division multiplexing (OFDM), which is recognized as the future of wireless communication solutions, has many merits such as robustness against multipath fading and inter-symbol interference (ISI), high spectrum efficiency, efficient implementation using fast Fourier transform (FFT) and easily adapt to severe channel conditions without complex time-domain equalization. However, OFDM system has some drawbacks. One of the drawbacks is the high peak to average power ratio (PAPR). When the amplitude of the signal exceeds the linear dynamic range of the amplifier, it will lead to signal distortion and spectral leakage which leads to a decline in the quality of communication.Among the techniques of OFDM PAPR reduction, the partial transmit sequence algorithm (PTS), which can reduce the OFDM PAPR efficiently without any distortion and limits of the number of OFDM sub-carriers, is most commonly used nowadays. However partial transmit sequence algorithm use the exhaustive search methods which results in high computational complexity and it is difficult to apply PTS algorithm into practical application.In this paper, we analyzed the principle of OFDM modulation and demodulation and the reasons that results in high PAPR in details. Based on the analysis, we proposed two improved PTS algorithms—PTS based on multi-loop search algorithm and a novel electromagnetism-like PTS algorithm.PTS based on multi-loop search algorithm utilizes multi-loop search method to instead the exhaustive search method in the traditional PTS algorithm, which could reduced the computational complexity effectively with almost no loss of PAPR performance.Compared to the traditional electromagnetism-like PTS algorithm, the novel electromagnetism-like PTS algorithm make a full use of particles'information which is reflected by the values of objective function to give a new charge definition to particles and we introduced a particle disturbance technique into the presented scheme to avoid premature convergence and improve the algorithm's ability of global searching. The simulation results show that the proposed method obtains more desirable PAPR reduction with lower computational complexity compared with the traditional EM-PTS scheme.