The Research Of Blind Equalizer Based On FPGA

In wireless communication systems, the multipath effect and the imperfections of the channel characteristics, lead to the bringing of iniersymbol interferenee when data transfer. In order to get the original source signal, an equalizer has to be used in the receiver to eliminate inter-symbol interference and compensate non-ideal characteristic of the channel. without the assistance of the transmitted sequence, The received data sequence is only needed by blind equalizer to reduce the ISI and equalize the channels. Blind equalization became a hot topic in recent years. Because of its computational complexity and good performance, this blind equalizer technology has a wide range of applications in the field of digital communications, wireless communications, image processing and radar.The article eirstly analyzes the fading characteristics of the wireless channel in mobile communication systems and the reason of inter-symbol interference generated, then introduces the equalizer technology and analyzes the several common blind equalization algorithm which are based on the the Bussgang nature, focusing on the constant modulus algorithm of the MATLAB simulation, researching the convergence rate and convergence error in the different channels, different balanced order, different noise ratio, different steps and channel mutations. Secondly, the FPGA design flow is introduced, in ISE9.2software environment, constant modulus algorithm of blind equalization system is achieved with a top-down digital system design using Verilog HDL language, It details the overall system block diagram, intriduces the function of the overall module and shows the block diagram and specific design of filter module, the judgment module, error module and the coefficient update module. The last, it first uses the ModelSim SE6.0test software on the design of blind equalization system simulation, and then in the Xilinx XUP Virtex-II Pro development board of equalizer simulation test shows the output waveform on an oscilloscope. In the comparison with the equalizer decision output waveform and the original source waveforms, we can see that there is error in equilibrium early, followed by exactly the same and the original source.From the simulation results,we can see that the waveform in the ModelSim SE6.0software simulation and the waveform on the oscilloscope are same, at the same time, it verifies the correctness of the design. The system can still achieve a balanced under the channel mutation, indicating that the design of the system has the ability to track channel. These fully proved that the blind equalizer in the error rate, convergence speed and channel tracking capabilities have good performance.