The Research On Frequency Offset Estimation In Dynamic Time-varying Channel With Low SNR

Compared with fixed-line,mobile communication technology allows people to communicate anytime,anywhere,greatly facilitating people's lives.However,with the advancement of society,the requirements for communication technology are getting higher and higher.For example,on a high-speed rail,the speed is as high as 400 km/h.Passengers may need to access the Internet for entertainment during a boring ride,even if they need to make a call during a business trip.At this time,the signal will suffer a large frequency offset,and the high-speed rail driving environment is also very complicated,and it may be surrounded by mountains,through caves,and so on.At present,the UAV technology is widely used in various occasions,such as large-scale aerial reconnaissance aircraft and civil aerial photography machines.These occasions require movement and even high?speed movement,and the surrounding environment is very complicated.Communication is often affected by mountains,clouds,and buildings.etc.There are many such examples.It has become an urgent need to communicate under low SNR and dynamic time-varying channels,but the motion must cause the frequency offset of the signal.Therefore,the research on low signal-to-noise ratio(SNR)and dynamic time-varying channel frequency shift estimation are very meaningful.Due to the Doppler effect,the difference in the transceiving oscillator,etc.,there is always a certain frequency offset between the transceiving signals.The frequency offset will have a great influence on the demodulation and descrambling of wireless communication,and it will cause inter-symbol interference in the OFDM system.Therefore,we need to accurately estimate the frequency offset of the received signal for later compensation.Eliminate the effects of frequency offsets.With the increasing demand for communication technology and low signal-to-noise ratio,frequency shift estimation in dynamic time-varying channels has become a problem.This paper compares and analyzes various frequency shift estimation methods,including time domain and frequency domain algorithms,data assisted and non-data assisted.The research algorithm can not achieve good results at the same time in accuracy and algorithm complexity.After comprehensive comparison,analysis and improvement,this paper proposes an algorithm integrating multiple FFT interpolation estimation methods.The simulation shows that the mean square error(MSE)of the frequency shift estimation of this algorithm is lower than 10-4 under the condition of SNR as low as 0 db,and will not cause a large fluctuation due to the change of the frequency offset value.The proposed frequency shift estimation algorithm is based on channel coefficients.At the same time,a channel estimation scheme in MIMO SC-FDE communication system is proposed,which can effectively resist the influence of Gaussian white noise and channel time-varying characteristics.Finally,the FPGA is used to implement the hardware implementation of the above channel estimation scheme and frequency shift estimation scheme.The simulation and field test show that the performance is good and meet the requirements of MIMO SC-FDE communication system.