Research And Application Of MIMO Detection Based On Semidefinite Programming

Without increasing transmitted power or expanding the bandwidth, the channel capacity and spectral efficiency can be improved efficiently in multi-input multi-output (MIMO) system.Because of the advantages, the MIMO technique has been widely used in the next generation of mobile communications and broadband wireless access systems.Vertical Bell Labs layered space-time (VBLAST) architecture can attain very high spectral efficiency with the application of spatial multiplexing. However, on the other hand, no diversity at the transmitter increases the bit error rate(BER) of system and then influences the communication quality. In order to solve the problem, detection algorithms with superior performance at the receiver are the key factor. In this thesis, we focus on the key technology of VBLAST architecture-MIMO signal detection technology. The main contents in this thesis are:1) This thesis introduces the basic principles of wireless channel characteristics in mobile communication and diversity technology. Then theoretically analyzes the system model and capacity at the environment of flat Rayleigh-fading channel, proving that the channel capacity is directly proportional to the minor number of transmit antennas and receive antennas in rich-scattering propagation environment.2) For the signal detection at the receiver, the basic principles, performance and computational complexity of various traditional detection algorithms, e.g. maximum likelihood detection (ML), linear detection (LD), nonlinear algorithm including successive interference cancellation without or with ordering (SIC or VBLAST) and sphere decoding (SD) are described in detail.3) The thesis puts forward two kinds of quasi-maximum likelihood algorithms based on semi-definite programming, SDR algorithm and PSK algorithm, which reduces the complexity of SD, while preserving the near optimal performance under various channel condition and signal to noise ratio (SNR) levels.