Equalization In Single Carrier Based On Fractional Fourier Transform

In the system of wireless broadband communication, the Inter Symbol Interference (ISI)may greatly affect the reliability of the communication performance. Single carrier frequencydomain equalization (SC-FDE) is an effective solution to ISI, this technology is utilized todesign a equalization coefficients to eliminate the interference caused by the channel and ISI.Currently available equalizations are to transfer the receiving end of the signals which sufferserious interference from time domain to frequency domain, and then design a frequencydomain equalization to equalize the signal. From the signal transfer point of view, thereceiving end of time-domain signal can be converted into a transform domain instead of thefrequency domain. and then the signal will be equalized in the transform domain. As a result ofthe transform domain is flexible, therefore by selecting the most matching channelcharacteristics the transform domain equalization can achieve better results than the frequencydomain equalization.Firstly, this paper introduces the development status of the SC-FDE technology as well asthe structure and characteristics. And then introduces the definition and nature of the fractionalfourier transform. The innovation point of this paper is that replacing the fourier transform inSC-FDE with fractional fourier transform. So as the signal can be equalized in the fractionaldomain, thus knowing about the definition and related nature of fractional fourier transform isvery necessary. And then, this paper focuses on some technologies about single carrierequalization based on fractional fourier transform. The MMSE equalization based onfractional fourier transform always use the least square error criterion.To calculate theequalization coefficient by minimizing the error. As the research executed in fractional domain,the calculated equalization coefficient contains fractional parameter. The innovation of thispaper is to achieve lower bit error rate than the frequency domain equalization by selection themost matching the characteristics parameter.The frequency domain equalization contains linear and nonlinear equalization. The decision feedback equalization, which is a nonlinear equalization, is worked as feeding backthe time-domain signal decided in the prior slot. And further eliminate ISI. This method is aglobal optimal method. Because the feedforward equalizer coefficients and feedback equalizercoefficients has some relations, the fractional-domain equalization can achieve optimalperformance from the whole by simultaneously adjusting the two coefficients and combiningthe advantage of fractional domain equalization. Theoretical analysis and simulation resultsshow that the new algorithm achieve lower BER than the frequency domain equalization.Finally, this article studies a new independent structure of the fractional domain decisionfeedback model on the basis of the decision feedback equalization based on fractional fouriertransform above. The model is to combine the feedforward equalized signal with the decisionfeedback signal, so as to eliminate the relation between the two parts. Thus the decisionfeedback equalization can be more flexible. After analyzing the theory, it is clear that theperformance of the new type is similar to the traditional one.