Investigation of MIMO-OFDM Peak-to-Average Power Ratio Reduction Techniques and the Implementation of the Corresponding Software-defined Radio Platform

For satisfying the growing demand of high data rates and good quality in wireless communications, IMT-Advanced (4G) seems to be a solution for future communication systems. This technology has a target transmission rate of 100 Mbps for low mobility users and 1 Gbps for high mobility users. Since two powerful techniques, orthogonal frequency division multiple access (OFDMA) and multiple input multiple output (MIMO) antennas, have been combined as a main transmission scheme in IMT-Advanced standard, the study of MIMO-OFDMA system is an important subject for the future world. In this project, we focus on peak-to-average power ratio (PAPR) reduction problems in the MIMO-OFDM systems. As we known, one major drawback of the OFDM technique is the high peak-to-average power ratio (PAPR) of the transmitter’s output signal. When an OFDM signal with high PAPR is passed through the nonlinear device, such as a transmit power amplifier, it may suffer from significant intermodulation distortion and out-of-band radiation. High PAPRs in OFDM signals also brings an increased complexity of the digital-to-analog and analog-to-digital converters and a reduced efficiency of the transmit power amplifier. Therefore, it is an important issue to develop a low-complexity and high-performance PAPR reduction method for MIMO-OFDM systems. In the first year of this project, we plan to use the extra degree of freedom, provided by the space-frequency block coded (SFBC) MIMO-OFDM systems, to develop PAPR reduction techniques. A preliminary method has been proposed in the 2009 IEEE Vehicular Technology Conference-Spring (VTC 2009-Spring). Compared with the conventional PAPR reduction method, this proposed scheme shows a better PAPR reduction performance without affecting the SFBC structure in MIMO-OFDM systems. Although this preliminary method has an exciting result, we find this method is not an optimal solution and still has plenty space to improve. Based on these observations, we aim to develop a general PAPR reduction method for SFBC MIMO-OFDM systems in this project and accomplish the corresponding floating-point and fixed-point simulations. We also note the problem of high computational complexity of the MIMO-OFDM PAPR reduction method and plan to apply our previous idea [91], proposed in the IEEE Transactions on Signal Processing, to develop a low-complexity PAPR reduction method for SFBC MIMO-OFDM systems. We also want to implement the developed low-complexity PAPR reduction algorithm on the Rice University's WARP (Wireless open-Access Research Platform), which is a scalable and extensible programmable wireless platform, built from the ground up, to prototype advanced wireless networks. Our laboratory already has two sets of WARP, and we expect to use the OFDM reference design provided by the open resource of WARP to establish a completed MIMO-OFDM transceiver system with multiple antennas and evaluate the transmission performance of the developed algorithms in this real system. We also expect the participated students in this project to have the ability to integrate the software and hardware design using the Xilinx System Generator and Simulink tools. We expect to accomplish the following items in two years: The First Year (August 2012 ~ July 2013): (1.1) Develop a systematic method to find all the possible SFBC coding patterns for MIMO-OFDM PAPR reduction. (1.2) Investigate an optimal method to combine the above SFBC coding patterns to achieve significant PAPR reduction. (1.3) Develop PAPR reduction algorithms for SFBC MIMO-OFDM systems and accomplish the corresponding floating-point and fixed-point simulations. (1.4) Establish a simplified MIMO-OFDM transceiver system using the WARP platform with the analog boards (RF circuits are not included). And let the participated students in this project to familiar with the co-design techniques for the Xilinx System Generator and Simulink tools. (1.5) Implement the developed PAPR reduction algorithm on the Platform developed in step (1.4) to verify the performance in real communication systems. The Second Year (August 2013 ~ July 2014): (2.1) Investigate the mathematical properties of the SFBC coding patterns found in step (1.1), finding the possible way to reduce the computational complexity of the encoding and decoding processes. (2.2) Utilize the method developed in [90] and [91] to reduce the computational complexity of the proposed SFBC MIMO-OFDM PAPR reduction algorithm (2.3) Develop a low-complexity and high-performance PAPR reduction algorithms for SFBC MIMO-OFDM systems and accomplish the corresponding floating-point and fixed-point simulations. (2.4) Use the radio boards of WARP platform to establish a MIMO-OFDM transceiver system with multiple antennas. And let the participated students in this project to have the ability to integrate the software and hardware design using the Xilinx System Generator and Simulink tools. (2.5) Integrate the system developed in step (1.5) and (2.4) to form a completed MIMO-OFDM transceiver system and evaluate the performance in real systems.

Project ID:PB10202-1068
External Project ID:NSC101-2221-E182-048-MY2