Design Of MIMO-OFDM System Based On GNU Radio And USRP

Orthogonal Frequency Division Multiplexing(OFDM)has become a research hotspot due to it's high spectrum utilization.Multiple input multiple output technology(MIMO)improves the data transmission rate of the system through multiple transceiver antennas.MIMO-OFDM technology has the advantages of both MIMO and OFDM technology,ensuring the spectrum utilization of the system while maintaining relatively high-speed transmission.In the MIMO-OFDM system,channel estimation and system synchronization are very important.Channel estimation can reflect the wireless channel conditions so that the signal can be demodulated correctly.In addition,the inevitable frequency offset in the wireless channel has a great impact on the orthogonality between OFDM subcarriers,and synchronization technology can effectively reduce this impact.This paper mainly studies the channel estimation and system synchronization technology in the MIMO-OFDM system.First,in order to further improve the timing synchronization accuracy of the OFDM system,this paper combines the sequence conjugation and sequence inversion methods in the three classic algorithms of Schmidl,Minn and Park.Moreover,improves the PN sequence which is commonly used in the classic algorithm to the ZC sequence.Theoretical simulation shows that the algorithm removes the peak part of the Schmidl algorithm,and only has one peak at the ideal position,which solves the problems of multiple peaks in the Minn algorithm and the sharp secondary peaks in the Park algorithm.The single peak of the improved algorithm is conducive to the detection of the peak position,and improve the accuracy of timing position.Secondly,for the problem that the inner product rule in the classical algorithm for channel estimation based on compressed sensing cannot filter out the best atom,this paper adopts an improved algorithm based on generalized orthogonal matching pursuit(GOMP).The improved algorithm replaces the classic inner product rule with the Dice atom matching criterion,and integrates the method of"weak"selection of atoms to avoid the problem that the inner product criterion cannot select the optimal atom from the redundant dictionary.Through Matlab theoretical simulation,when the step size is set to 3 and the signal-to-noise ratio is 40d B,the error rate of the GOMP algorithm is 0.8557×10^-2,and the error rate of the improved GOMP algorithm is0.2356×10^-2,which proves the improved GOMP algorithm is better than GOMP algorithm in bit error rate.Finally,this paper builds a MIMO-OFDM wireless communication system based on GNU Radio software and USRP(Universal Software Radio Peripheral),and applies the improved channel estimation algorithm and synchronization algorithm to the system.Experiments show that the system can successfully send and receive images and videos,in order to further compare and verify the transmission and reception performance of the system,the transmission and reception test of different texts is carried out.Experiments show that the improved algorithm performs better than the classic algorithm in receiving and sending in the actual channel.