Research And Implementation On Blind Equalization Techniques For High Order QAM

Compared to low order modulation,higher order QAM modulation is more and more widely used in high-speed broadband communication systems due to its high band efficiency.During the high-speed transmission of digital signal over the actual channel,intersymbol interference introduced by the bandwidth-limited and multipath effect,will cause the distortion of the received signal in both amplitude and phase,resulting in larger error rate for symbol detection at the receiver,thus a equalizer is necessary for the received signal.In the scenario where the training sequence can not be provided,blind equalization technique must be adopted.In the vector signal analysis system,with the increase of QAM modulation order and the transmission symbol rate,intersymbol interference introduced by the channel will affect the received signal badly,and at the same time,complexity of blind equalizer in the receiver will also increase.Therefore,it is of great significance to study blind equalization algorithms for high-order QAM for vector signal analysis systems.In order to implement blind equalization for 64 QAM and 256 QAM in vector signal analysis system,this paper studies two conditions wether there exist carrier frequency offset.Firstly,the equivalent baseband model of digital modulation and demodulation system is established,which lays the foundation for further research.Next,without considering the existence of carrier frequency offset,the multi-modulus blind equalization algorithm for high order QAM modulation and its improved algorithms are analyzed and simulated,a dual mode blind equalization algorithm based on SMMA is finally adopted as blind equalization scheme of high order QAM for the condition without frequency offset.Then,according to the situation with large frequency offset,based on the analysis of influence of carrier frequency offset and phase offset on the blind equalization algorithms,and research on several kinds of closed-loop carrier recovery algorithms,two kinds of structures of joint blind equalization and carrier recovery are given.The simulation results show that the joint structure II can achieve convergence with less symbols than the joint structure I does,however,it takes plenty of time for the CMA pre-equalizer to reach the steady state.After that,the scheme of joint blind equalization and carrier recovery is defined for vector signal analysis system.Finally,an implementation framework of joint blind equalization and carrier recovery is given for high-order QAM at high symbol rate in vector signal analysis system.The theoretical analysis and the actual data test show that the scheme introduces little error which can be ignored.And besides eliminating ISI at the actual symbol rate of 300 Msps,it can correct a frequency offset range of 10% of the symbol rate,and under the three test conditions with different frequency offset,the output EVMs have all decreased to around 1%,which verifies its effectiveness.