Research On Peak-to-Average Power Ratio Reduction Algorithms For Orthogonal Frequency Division Multiplexing Systems

Orthogonal frequency division multiplexing(OFDM)technology is one of the key technologies for terrestrial wireless communication systems.And it is also a promising technology for satellite mobile communications,next-generation global navigation satellite system,and remote sensing satellite systems.However,OFDM suffers from the problem of high peak-to-average power ratio(PAPR),which will cause non-linear components in the communication system,especially RF power amplifiers,to have higher nonlinear distortion or lower power efficiency.In order to improve the linearity and efficiency of the communication systems,it is necessary to reduce PAPR of OFDM signals.Numbers of studies have been carried out to work on PAPR reduction,but existing works have not fully considered the tradeoff between the level of PAPR reduction and the system performance,nor the involved high computational complexity.Compared to those traditional PAPR reduction algorithms,this dissertation comprehensively studies the PAPR problem and provide the proposed solutions considering PAPR reduction,bit error rate(BER),out-of-band distortion,and computational complexity.Then several algorithms were designed and related simulations have been carried out to evaluate their performances.The main contents of this research are listed as follows.The metrics of envelope dynamics of OFDM signals are evaluated.The metrics,i.e.,PAPR,cubic metric,and the recently proposed distortion component measurement,were tested and evaluated in terms of the prediction of power back-off.It is verified that these metrics are consistent with each other on the description of envelope dynamic.Therefore,the PAPR,which has the lowest computational complexity,was selected for the signal processing on envelope dynamic suppression.A clipping-noise compression based algorithm for peak-to-average ratio reduction and its further improved version are proposed.In the traditional clipping-based algorithms,the clipped signals usually need to be filtered to meet the requirements of the transmit spectral mask.However,the post-clipping operations will cause signal peaks to regrow,leading to the decreasing of the PAPR reduction.In order to address this problem,this dissertation proposes a novel algorithm named clipping noise compression,and its improved version.The proposed algorithm suppresses the regrowth of peaks by compressing the amplitude of the clipping noise.Then the improved algorithm reduces the computational complexity of the signal processing by introducing a preset constant normalization factor.The simulation results show that compared to the traditional clipping-based methods,both of the proposed algorithms achieved improved performance on PAPR reduction and BER.A novel iterative algorithm for PAPR reduction with low computational complexity is proposed.The traditional iterative clipping and filtering technology can address the problem of peak regrowth through iterations.However,the iterative operations usually result in the latency of signal processing and high computational complexity.To address this problem,this dissertation proposes a low complexity algorithm of iterative clipping.In the proposed algorithm,the clipping noise compression technique is introduced to replace the frequency-domain filtering,resulting in lower computational complexity.The algorithm also improves the calculation efficiency of the clipping threshold.The simulation results show that compared with traditional algorithms of iterative clipping and filtering,the proposed algorithm achieves better performance on PAPR reduction and BER.Two algorithms on PAPR reduction are proposed,which jointly use clipping and companding methods.Both of the two algorithms carry out signal processing merely in the transmitter and do not require signal recovery in the receiver,leading to the reduction of the complexity for system design.In the first algorithm,i.e.,the clipping based joint algorithm,the amplitudes of the time-domain signal are clipped to a threshold,and the obtained clipping noise is companded and used to modify the clipped signal to obtain the transmmited signal.In the second algorithm,i.e.,the companding based joint algorithm,the clipping threshold is introduced for the companding function,which is used to compress the large signal while maintaining the small signal unchanged.The simulation results show that both of the joint algorithms have the advantages of PAPR reduction and BER comparing to the traditional clipping technology or companding technology.