Research On Adaptive Transmission Technique For MIMO-SCFDE System

As wireless communications have been experiencing the rapid development, researchers have been highly concerned about how to increase the date transmission rate, to cope with the growing need for communication services. Up to date, multiple-input multiple-output (MIMO) and orthogonal frequency division multiplexing (OFDM) are regarded as two effective methods to obtain higher spectral efficiency. Because MIMO and OFDM are the spatial and frequency domain approaches respectively, we can combine OFDM with MIMO to improve the spectral efficiency further.Firstly, this thesis gives a brief introduction to block-wise transmission and MIMO technique, and analyzes how the cyclic prefix (CP) converts the linear convolution into circular convolution. Furthermore, different receivers for MIMO system, antenna selection and precoding technique are discussed.When the transmitter has the knowledge of the channel state information (CSI), adaptive technique can exploit the channel better by changing the transmit strategy adaptively, to improve the data transmission rate or the system performance. In this thesis, we summarize the exiting adaptive schemes for single-input single-output (SISO) block-wise transmission and MIMO-OFDM systems, and pay attention to the latter one. For adaptive MIMO-OFDM systems, the transmitter performs some signal processing based on CSI, e.g. optimally allocating the power across frequency and space, determining the optimal precoding scheme or beamformer, etc. However, the performance improvement of adaptive MIMO-OFDM is usually at the cost of large amount of CSI feedback. What's more, some adaptive algorithms for MIMO-OFDM systems involve high complexity.Compared to OFDM, single carrier with frequency domain equalization (SC-FDE) has the same immunity to multipath fading and almost the same complexity and spectral efficiency. But SC-FDE eliminates the Peak-to-average Power Ratio (PAPR) problem. For SISO system, adaptive SC-FDE system based on optimal signal subspace (OSS) has been proposed, which requires only one bit feedback per frequency sub-channel and obtains simple BER controllability. Therefore, we hope this adaptive SISO SC-FDE can be extended to MIMO system, while the system proposed can maintain low feedback and simple BER controllability.This thesis proposes a low-cost adaptive MIMO-SCFDE system which performs frequency sub-channel selection, and gives the definition of OSS for MIMO system and SNR criterion to determine the frequency sub-channels used to transmit data symbols. Furthermore, considering the fact that rank-deficiency and ill-conditioning of the MIMO channel would discard all good spatial sub-channels of some frequency sub-channel, we introduce precoding technique into this adaptive MIMO-SCFDE system. We employ the same simple precoding matrix on the frequency sub-channels used, and therefore the adaptive MIMO-SCFDE system combining precoding and frequency sub-channel selection needs nearly one bit CSI feedback per frequency sub-channel. The simulation results show that the adaptive MIMO-SCFDE system employing precoding and frequency sub-channel selection can considerably improve the transmission rate than without precoding. In addition, the simulation compares the adaptive MIMO-SCFDE system with interpolation-based multi-mode precoding MIMO-OFDM system, and shows that when fixing the data rate and average transmit power the system we propose have obvious advantage of feedback overhead over the latter.