Research On Differential Detecting Of Quasi-Orthogonal Space-time Code And Its DSP Implementation

MIMO (Multi-input Multi-output) system can improve channel capacity of wireless communications without increasing the transmit power or spreading the system bandwidth, so it attracts the most attention and is studied widely. Space-time coding which can provide space diversity is an implementation of the MIMO technique. STBC (Space-Time Block Codes) becomes one of most important Space-time codes, because it has some advantages such as the simple construction, low decoding complexity and good performance.Most researches on STBC are assumed that the CSI is available at the receiver. However, it is difficult to estimate CSI, or it will increase the complexity and cost of system obviously in which the receiver is in a quickly moving environment, so that the differential space-time codes is put forward, it doesn't need the CSI and can be decoded by the adjacent receiver signals. But decoding complexity of the existing differential detection is exponential to the transmit rate, and the good performance is obtained only at high SNR (signal to noise ratio) because of the constellation expansion. So we study the differential detection of quasi STBC, the main contents include:1. Based on the noncoherent unitary space-time modulation, we introduce a differential detector with lower complexity for MIMO system, and ML (Maximum Likelihood) search space size is linear to modulation rate at receiver, to avoid the constellation expansion. Then we combine it with OFDM (Orthogonal Frequency Division Multiplexing), and carry out differential detection for MIMO-OFDM. The proposed differential scheme has lower BER (Bit-Error-Rate) than that of the differential detection with full-diversity at lower SNR (Signal Noise Ratio), although it's partial transmit diversity.2. Then we carry out the codec with C and assembly language on the basis of TMS320C5416 DSP. The RTDX debugging result shows that bit error rate computed by DSP is similar to that of the MATLAB's, and the simplified proposed algorithm can reduce the processing time of decoding by using DSP implementation.