Low PAPR Linear Precoding

Massive multiple-input and multiple-output(MIMO) system has received widely attention from all over the world since it can greatly improve the energy efficiency(EE) and spectrum efficiency(SE) of the wireless communication system. While in a frequency selective fading channel, channel delay spread in time domain will cause severe inter-symbol interference(ISI), leading to the increase of bit error rate(BER). Thus, the orthogonal frequency division multiplexing(OFDM) technology is often applied to the massive MIMO system to resist frequency selective fading. However, OFDM based massive MIMO system has a problem of high peak-to-average power ratio(PAPR) which can result in the nonlinear distortion and extra energy cost of the power amplifier. Therefore, we must take effective measures to reduce the PAPR.The pre-coding technique is an effective and flexible way to reduce the PAPR of OFDM signals. However, pre-coding schemes usually increase error probabilities of the system. In my thesis, a pre-coding scheme is introduced and the procedure to design such a pre-coding matrix is also provided. The performances analyses of uplink MIMO-OFDM system are provided for both single user and multiple users scenarios. For single user scenario, we analyze the system BER performance and derive a necessary condition on the chosen pre-coding matrices for minimizing error probability of the signals with the knowledge of the channel information and receiver filter. Besides, the EE and SE performances of the pre-coded system are analyzed as well; For multiple users scenario, the theoretical signal to noise rations while the number of receiver antennas goes to infinite are derived. Numerical results show that the optimal pre-coding matrix could reduce the PAPR without losing BER performance for both scenarios. Moreover, in Raleigh multi-path channels, system BER and EE performances of the pre-coded system are increased compared with the system without pre-coding for single user scenario.