Research On Frequency Offset Estimation Algorithm Of MIMO-OFDM Systems In Doubly Selective Fading Channels

Orthogonal frequency division multiplexing(OFDM)technology combined with multiple input multiple output(MIMO)technology can effectively combat multipath fading,increase channel capacity and improve spectrum efficiency,which has become one of the key technologies of broadband mobile communication systems.In OFDM-based systems,carrier frequency offset(CFO)results in inter carrier interference(ICI),which significantly deteriorates the performance of MIMO-OFDM systems.Therefore,CFO estimation has important theoretical significance and practical application value.When MIMO-OFDM system works in doubly selective fading channels,the traditional CFO estimation algorithms can't overcome the influence of ICI introduced by doppler spread on the received signal,and therefore suffers from a certain degree of performance loss.In this paper,the CFO estimation algorithm for MIMO-OFDM systems in doubly selective fading channels is studied.The pilot-based frequency offset estimation algorithm and the banded structure algorithm are improved by ICI self-cancelling precoding technology.The improved algorithm effectively reduces the ICI introduced by doppler spread in the received signal,thus improving the performance of frequency offset estimation.Firstly,the basic principle of MIMO-OFDM is introduced.The influence of CFO on MIMO-OFDM systems and the influence of doubly selective fading channels on CFO estimation are analyzed.In this paper,an improved Moose estimation algorithm based on pilot assisted is studied,and it is extended to MIMO-OFDM systems.The extended algorithm can achieve satisfactory performance in static channel,but the performance loss is serious in doubly selective fading channels,and it can only estimate decimal CFO,which limits its application in practical systems.Secondly,this paper studies the maximum likelihood(ML)algorithm and the banded structure algorithm.The maximum likelihood estimation algorithm based on the base expansion model reduces the number of parameters to be estimated.However,the performance loss of this algorithm in the doubly selective fading channels is large.The banded structure algorithm is simple in principle,and has low computational complexity.However,with the increase of signal-to-noise ratio,the estimation performance of the algorithm tends to be fixed gradually.And,neither of the two algorithms can overcome the influence of ICI introduced by doppler spread on CFO estimation,so they can't be applied in doubly selective fading channels.In view of the fact that the above algorithms can't overcome the ICI introduced by doppler spread,the following two aspects are studied in this paper.Firstly,an improved pilot estimation algorithm based on ICI self-cancellation precoding is proposed.The basic principle of ICI self-cancelling precoding is analyzed in detail.It is shown that ICI self-cancelling precoding can effectively reduce the influence of ICI introduced by doppler spread on received signals,thus improving the performance of CFO estimation.The pilot structure is reconstructed by self-cancelling precoding modulation technology at the transmitter,and the received signal is obtained by self-cancelling precoding demodulation technology at the receiver.Then the CFO estimation of the improved algorithm is obtained by correlation.Simulaton results show that the performance of the improved algorithm is improved by ICI self-cancelling precoding,and the estimation performance is more stable under different CFO values.Secondly,the banded structure algorithm is improved by using ICI self-cancellation precoding.The improved algorithm uses ICI self-cancelling precoding modulation to construct the transmitting signal,constructs the covariance matrix at the receiving end based on the new received signal vector,and estimates the CFO by the band structure property of the covariance matrix.Simulation results show that ICI self-cancellation precoding can reduce the influence of ICI introduced by doppler spread on received signal,thus improving the performance of the banded structure algorithm.