Sub Channel Joint Precoding Technology In OFDM Systems

For wireless channel with frequency selective fading,Orthogonal Frequency Division Multiplexing(OFDM)is a popular technology in the next generation communication systems,which can divide the wireless channel into multiple flat fading sub channels,improve the spectrum utilization,and enhance the ability to resist multi-path fading.In OFDM system,because of the non ideal characteristics of wireless channel,it may lead to a serious decline of some sub carriers in the transmission process.If there is no preprocessing in transmitter,the data loss caused by the sub carrier fading in OFDM system can not be overcome.The precoding technology can solve the problem of data loss caused by sub carrier fading.In OFDM,since the precoding module is in front of the IFFT,it is also known as the sub channel joint precoding technology.This paper mainly did research on the channel independent precoding.From the point of maximum likelihood detection algorithm,the accuracy of the pairwise error probability(PEP)formula was derived,combined with the statistical properties of the modulation constellation,this paper proposed an independent channel information precoding scheme.Simulation results showed that,for the specific modulation,the proposed precoding scheme outperformed the currently existing channel independent precoding schemes.In this paper,we also have studied the signal detection model of the OFDM system with precoding.we proved that,at the receiver,the zero forcing equalization(ZF)algorithm was not suitable for independent channel information precoding OFDM systems.Moreover,combined with the nonlinear detection algorithm,the precoding technology can bring a substantial increase in performance.Further analysis proved that the nonlinear signal detection of the OFDM system with precoding is equivalent to multiple input multiple output(MIMO)signal detection.Under the condition that the probability of the transmitted symbol is equal,the optimal algorithm for MIMO signal detection is the maximum likelihood detection(ML)criterion,but the complexity of ML detection increases exponentially with the number of transmit antennas and the size of the constellation of MIMO,so a series of improved algorithms with low complexity have been proposed,such as spherical decoding(SD)algorithm,the maximum likelihood decoding based on QR decomposition and branch search(QRM-MLD)algorithm,etc..In this paper,we first carried on the mathematical model of the modulation constellation,and then realized the ML detection algorithm by using the penalty function method with low complexity;we also studied the QRM-MLD algorithm by introducing the constraint condition of noise statistics,the simulation results showed that the complexity of the improved QRM-MLD algorithm can be further reduced without affecting the detection performance.