Research On New CFO Estimation Method And Its Application

With the development of wireless communication,the market has higher and higher requirements for communication quality.Faster transmission rates,larger capacity,and more efficient spectrum utilization have become the research direction and goal of non-mathematicians.Orthogonal Frequency Division Multiplexing(OFDM)technology emerges as a multi-carrier technology.Its greatest feature is that it can overcome the problem of frequency selective fading caused by multipath effects.High-speed serial data streams are converted to low-speed parallel data streams,which greatly improves spectrum utilization.Multi-Input-Multi-Output(MIMO)technology is a multi-antenna technology that can further improve the performance of communication systems.On the basis of traditional MIMO,the number of larger-scale antennas deployed on the base station side has evolved into the currently popular massive MIMO technology.Since massive MIMO has poor performance under frequency-selective fading channels,and OFDM technology can just overcome frequencyselective fading,the combination of the two technologies can maximize the advantages of the two technologies.However,in order to obtain the most excellent performance requires effective acquisition of channel state information.At the same time,massive MIMO-OFDM technology also inherits the disadvantage that OFDM is sensitive to Carrier Frequency Offset(CFO),which undoubtedly seriously affects system performance.Therefore,effective frequency offset and channel estimation are the keys to improving the performance of massive MIMO-OFDM systems.Most of the traditional frequency offset and channel estimation are analyzed in time-invariant scenarios.However,this is not suitable for time-varying channel scenarios.There are too many parameters to be estimated in time-varying channels.Basis Expansion Model(BEM)can solve this problem,it can simplify the quantity to be estimated and become the first choice for parameter estimation in time-varying channel scenarios.Especially with the development of high-speed rail,channel and frequency offset estimation under fast time-varying channels has become the key.CFOs often interact with non-ideal channels,so combining the two for joint estimation can further improve system performance.The core work of this paper is to study the joint frequency and channel estimation algorithms for time-invariant and time-varying channels.For time-invariant channel scenarios,the proposed algorithm is based on the Maximum Likelihood Estimation(ML)algorithm and uses the ZC sequence as a training sequence to reduce the computational complexity.Also,it has a lower peak-to-average ratio(PARR)for the system.For time-varying channel scenarios,in order to reduce the increased number of estimated parameters in the time-varying channel,a time-varying channel is modeled using a base extension model.The maximum posterior(MAP)estimation algorithm is used to for the joint estimation.Simulation results show that the frequency offset and channel estimation in both scenarios have good performance.