Research And Implementation Of Downlink Channel Estimation Algorithm For Narrow-band Internet Of Things

The narrow-band Internet of things focuses on the low-power and wide-cover Internet of things market.As a new Internet of things technology,it has become an important branch of the interconnection of all things.The channel transmitted by narrowband Internet of things belongs to wireless channel,and the signal will be disturbed in the process of wireless channel transmission.Channel estimation can provide channel condition to correct the received signal and improve the system performance.Therefore,channel estimation is an indispensable link in narrowband Internet of things system.This paper focuses on the channel estimation technology of narrow-band Internet of things downlink.This paper first summarizes the narrowband physical network system,then studies the wireless channel transmitted by the NB-IoT,and establishes the NB-IoT channel model as Rayleigh fading channel.According to the three Rayleigh fading channel models specified in the 3 GPP protocol,the wireless channel simulation model is established as the EVA channel.Focus on the NB-IoT downlink channel estimation based on pilot,the main work of this paper is as follows:The pilot distribution scheme of narrowband Internet of things single antenna port is studied.Because the downlink pilot reference signal is NRS densely distributed in the time domain,the time domain distribution interval between the two adjacent pilots of the same subcarrier is large,which is not conducive to accurately calculating the channel parameters of the data points between the pilots.Based on Nyquist sampling theorem of pilot interval design principle,the time domain interval of preset pilot is improved.By increasing the number of pilot signals reasonably and shortening the time domain interval distance of the preset pilot,the pilot signal is distributed more dispersed in the time domain.The simulation results show that the improved pilot distribution scheme reduces the frequency band utilization of the system,but the bit error rate is obviously smaller than the preset pilot,which significantly improves the channel estimation performance of the system.The channel estimation algorithm based on pilot in narrowband Internet of things is studied.DFT algorithm is used as the channel estimation algorithm for the NB-IoT downlink system after analyzing the characteristics of the common pilot point channel estimation LS algorithm,the DFT algorithm and the LMMSE algorithm,and the threshold threshold threshold for filtering noise in the cyclic prefix in the time domain impact response(CIR)is improved by adding a Cassette window to suppress energy leakage in the non-integer sampling interval channel,The simulation results show that the threshold threshold can filter the noise better.The simulation results of DFT improved algorithm and other channel estimation algorithms show that the energy leakage of DFT improved algorithm is not obvious in the non-integer sampling interval EVA channel,and the new threshold filter is used to remove more noise.The performance of DFT algorithm is improved significantly.The original LTE framework is modified according to the NB-IoT system characteristics in the DLSIM downlink simulation link of the software radio platform based on GPP,and the platform architecture is redesigned.The physical downlink control channel NPDCCH and the physical downlink shared channel are built NPDSCH,and the process of signal processing is described.The data position interpolation algorithm of the narrowband Internet of things downlink system is studied.Because the pilot signal NRS the characteristic of two-dimensional discrete distribution of time frequency,it needs to be interpolated in frequency domain and time domain respectively.The interpolation in frequency domain can be realized by DFT the Fourier transform characteristic of channel estimation algorithm,which is equivalent to frequency domain interpolation.After analyzing the characteristics of data interpolation algorithm,constant interpolation algorithm,linear interpolation algorithm and Gao Si interpolation algorithm,the simulation results and computational complexity are obtained,The linear interpolation and constant interpolation algorithms are selected as the interpolation algorithms at NB-IoT data according to the characteristics of NB-IoT system and pilot distribution.The platform architecture is designed on the basis of the NB-IoT system characteristics in the DLSIM downlink simulation link of the GPP software radio platform.The physical downlink control channel NPDCCH and the physical downlink shared channel are built NPDSCH,Because the pilot reference signal is sent in the NPDSCH subframe,the improved pilot distribution scheme and process.