Combination Of Signal Processing In Spcae,time And Frequency Domain To Reduce The PAPR Of MIMO-OFDM Systems

Because it can achieve high spectral efficiency and effective against multipath fading, orthogonal frequency division multiplexing(OFDM) technology has become one of the key technologies for 4G communication systems. However, high peak to average power ratio(PAPR) is one of disadvantages of the OFDM signals. Signals with high PAPR require the non-linear devices such as power amplifier(PA) having large linear range. To transmit the signal which exceeds the linear range of the device will lead to nonlinear distortion and the bit error rate(BER) performance degradation.Multiple input multiple output(MIMO) technology makes full use of the space dimension. Then, it can improve channel capacity without adding frequency spectrum resource or sending power of antenna. The MIMO-OFDM systems also have PAPR problem for each antenna transmits the OFDM signals and the PAPR problem will be more serious with the increasing of antenna number. Compared to the mature PAPR reduction method in OFDM systems, there are few references to study the PAPR problem in MIMO-OFDM systems. The PAPR reduction method in OFDM, especially in Alamouti MIMO-OFDM will be further researched in this paper.PAPR reduction method in OFDM systems can be roughly divided into three categories: pre-distortion, coding and probabilistic. In the third category: selective mapping(SLM) and partial transmit sequence(PTS) are the promissing PAPR reduction method with good PAPR reduction performance and not causing additional nonlinear distortion. However, the conventional SLM and PTS method have two drawbacks.Firstly, they have high computational complexity for using too much Inverse Fast Fourier Transform(IFFT) oprerations. Secondly, in order to recover the original signal,the index of the selected rotation sequences for each antenna will be transmitted as side information(SI) by using extra subcarriers.In this paper, several modified methods have been proposed based on the conventional SLM and PTS in time domain and frequency domain, respectively.Multi-methods were combined to reduce the computational complexity and SI, even realize the signal blind detection. The main work of this paper is as follows:(1) A modified PAPR reduction method base on subclass signals swapping in OFDM systems. The candidates of conventional SLM will be divided into four subclasssignals by using IFFT properties in time domain. More candidates will be generated by four subclass signals swapping and synthesising. Compared with the conventional SLM,the signal process of the proposed method takes place in time domain and few complex addition operations will be used to generate more candidates. So, the computational complexity will be further reduced.(2) PAPR reduction for MIMO-OFDM systems in time domain. At the transmitting part, more time domain candidates will be generated by combination of the the signals’ cyclic shift, addition-subtraction between the antennas signals, sub-blocks exchange and equivalent coding, etc. The computational complexity of MIMO-OFDM systems will be futher reduced. At the receiving part, the property of FFT operation(time domain cyclic shift for OFDM signal corresponding to the frequency domain signal phase rotation)will be used. The reverse rotation signal will be obtained after FFT demodulation. To calculate the distances between reverse rotation signal and its’ nearest constellation points, and choose phase rotation factor corresponding to the minimum distance to restore the original signals.(3) PAPR reduction for SFBC MIMO-OFDM systems in frequency domain. At the transmitting part, the phase rotation factor for SLM and the partition method for PTS will be specially designed. Then, it can use the phase rotation factor individually for each antenna of the SFBC MIMO-OFDM systems and the coding matrix is also orthogonal. Each antenna can selected the transmitting signals with minimum PAPR from all the candidates so that the PAPR reduction performance is improved. At the receiving part, a suboptimal maximum likelihood(ML) detection method is propoed to recover the original signals. Each two adjacent subcarriers will be devided into a group for detection. The product of the phase rotation factors between the antennas will be detected and the original signals will be recovered by the detected phase rotation factors.Then, the blind detection of the received signal will be realized.