| Blind equalization technique is one of key technology in communication system. Under the premise of no increasing the bandwidth of the system, it can improve bandwidth utilization rate, and avoid inter-symbol interference (ISI) introduced by limited bandwidth and multi-path transmission. Because PSK and QPSK signal are widely used as modulated signal in the communication system, it uses QAM modulated signals of high bandwidth utilization rate and low bit error rate as input of equalizer. Combine with wavelet,multiwavelet theory and fractionally space, it proposes several channel equalization method based on coordinate transformation. And make simulation to algorithm use Matlab, analysis the performance of various algorithms. At the same time, it simulates two algorithms using TMS320VC5502 chip, give the software design process and make detailed description to the main module of the program. The main research contents are as follows:1. Aiming at high-order Quadrature Amplitude Modulation signals (QAM) are widely used in communication systems, it research algorithm to make equalization to 16QAM signals. Because 16QAM signals distribute in a few known semi-diameter circle, make use of error function of Constant Modulus Algorithm (CMA) to adjust weight vector, it leds to slow convergence and large mean square error. So, Constant Modulus Algorithm based on coordinate transformation (CTCMA) is proposed. The algorithm makes coordinate transformation to output signal of equalizer, error function of CMA obtained will become exactly zero when the channel is perfectly equalized. 16QAM signal are equalized well. Simulation results with an underwater acoustic channel model demonstrate that the algorithm has superiority, and the various parameters are obtained.2. The wavelet and multi-wavelet theory are introduced to blind equalization, the orthogonal wavelet Constant Modulus Algorithm based on Coordinate Transformation (CT-WTCMA) and the balanced orthogonal Multi-Wavelet Constant Modulus Algorithm based on Coordinate Transformation (CT-MWTCMA) are proposed. Input signal of equalizer are transformed using different wavelet, it reduces relevance of the signal and speeds up the convergence speed. Simulation results with an underwater acoustic channel model demonstrate that the two algorithms have faster convergence rate and smaller steady-state error, compared with CMA and CTCMA.3. Aiming at higher order QAM, T/2 Fractionally Spaced Constant Modulus Algorithm (T/2-FSE-CMA) has disadvantages of low convergence speed and large steady error, a Modified T/2 Fractionally Spaced Coordinate Transformation Constant Modulus Algorithm (T/2-FSE-MCTCMA) is proposed. In the proposed algorithm, real and imaginary parts of input signal of T/2 Fractionally Spaced Blind Equalizer (T/2-FSE-CMA) are separated to carry out equalization, and output signals of equalizer are transformed to the same unit circle by making coordinate transformation, a new error function of CMA is obtained after making coordinate transformation. The new error function is made to adjust weight vector of each sub-channel in T/2 fractionally spaced blind equalizer. The algorithm overcomes large misjudgments in using T/2 fractionally spaced blind equalization algorithms for equilibrating to higher order QAM of multi-modulus characteristics. Simulation results with an underwater acoustic channel model demonstrate that the proposed algorithm has faster convergence rate and lower residual error, compared with T/2-FSE-CTCMA and T/2-FSE-CMA.4. According to simulation parameters from Matlab, CTCMA algorithm and T/2-FSE-MCTCMA algorithm are simulated by using TMS320VC5502 chip. CCS is used as development environment of DSP. In the CCS, code of the algorithm is compiled using C language, and is optimized. Eye diagram of signal is showed on the host successfully. It has some reference value to design equalizer in communication system of 16QAM.Figure [39] Table [4] Reference [79]... |
PDF Full Text Request