Measurement-based Research On The Massive MIMO Channel Propagation Characterics And Modeling

As a hot spot of current research,Fifth Generation(5G)New Radio(NR)technology is expected to meet the explosive growth of mobile data traffic due to the advantages of higher transmission rate,lower latency and lower energy consumption compared with existing communication technologies.Massive multi-input and multiple-output(MIMO)can fully exploits spatial degrees of freedom by configuring tens or even hundreds of antennas at the base station to improve the spectral efficiency,the multi-user multiplexing and inter-cell interferencem,which is regarded as one of the key technologies of 5G NR.In order to better evaluate Massive MIMO technology,we carried out external field measurements and theoretical analyses in multiple scenarios.The main contributions of this dissertation are concluded as follows:1.Massive MIMO channel measurement,data post-processing and the rationality verification.We conducted extensive field measurements by the MIMO channel measurement platform(Sounder)and dedicated an-tenna arrays.In measurements,the 64,128,and 256 element transmition antennas are virtual spliced by using an orthogonally polarized rectangu-lar antenna array with 32 elements,and 16 elements in a orthogonally-polarized cylindrical array are used for reception antenna.In addition,the measure are conducted in various scenarios inlcuding typical urban mi-crocells(UMi),urban macrocells(UMa),outdoor to indoors(02i)with measuring carrier frequencies of 3.5 GHz and 6 GHz,and the system bandwidth of 200 MHz.Through data post-processing,we obtain channel impulse response(CIR)and extract seven-dimensional channel propaga-tion parameters.In addition,the rationality of the splicing measurement scheme is verified,then the reliability of subsequent parameter analysis is guaranteed.2,Analysis of Favorable Propagation Characteristics(FPC)of Massive MIMO based on measurement.In order to study the FPC of Massive MIMO,with the CIR and the channel propagation parameters,the statis-tical distribution characteristics in multiple scenarios with multiple fre-quency and different antenna numbers are compared from the delay dis-tribution,angular distribution,eigenvalue spread and channel capacity,it is concluded that:As the antenna elements increases,the consistency of transmission characteristics among the time domain does not change sig-nificantly,which means that the parameters related to the delay domain of 5G NR does not need to change greatly.But the power angular spectrum of Massive MIMO becomes more dispersive,and angle spread becomes larger in the angle domain.Furthermore,with the increase of antenna ele-ments and the carrier frequencies,the signals are more evenly distributed,and tend to be orthogonal.Thereby,more degrees of freedom are pro-vided,which leads to a remarkable improvement of channel capacity,e.g.,in the 021 scenario,when the transmitter elements are increased from 32 to 256,the multi-user MIMO capacity at the 3.5 and 6 GHz are increase by 54%and 48%,respectively.3.Performance analysis of Massive MIMO systems in terms of the spa-tial structure and elements number.In order to investigate the tradeoff between the limited size in small cell network and the capacity gain from increasing antenna elements,the spatial performances of Massive MIMO are studied.Using the data extracted from the 512 × 16 CIR matrix,the capacity of 32,64 and 128 elements with 0.5,1 and 2 wavelength(totally 9 cases)are compared.The results show that with the increase of an-tenna element spacing,the channel will tend to be low correlation,and the capacity gap are obtained,but which decrease to a certain degrees when the antenna elements are increased,e.g.,when the transmitter elements increase from 32 to 128,the capacity gap of antennas with element spac-ing of 1 and 0.5 wavelength are decreased from 19.3%to 7.6%.This is caused by the interaction between the capacity gain from elements spacing and the gain from the elements number.Which imply that when design-ing a Massive MIMO antenna with huge amounts of antenna elements,a smaller antenna element spacing should be taken into consideration.4.Investigate the required antennas number at the transmitter of Mas-sive MIMO system based on Zero-forcing(ZF)precoding.The effect of finite number of BS antennas on the performance capability of ZF pre-coder in a rich scattering channel is quantified through theoretical deriva-tion and actual measurement.Two insights have derived from the theo-retical analysis:the needed number of the BS antennas for ZF precoder with the optimal power allocation to achieve certain percentage of the BC capacity is monotone decreasing when the transmit signal-to-noise ra-tio(SNR)is increased and an upper bound of the needed transmit antennas is derived with simple expression which is only related to the ultimate percentage and the number of active users.The theoretical derivation are verified through simulation and the performance of actual measurements are compared.In the actual channel,when the correlations between the antenna arrays are weak,the measured results are consistent with the the-oretical derivation.When the correlation between the array elements is strong,e.g.,when the antenna array spacing is less than 1 wavelength at 3.5 GHz,the ZF precoding performance is lower than the derivation value at low SNR(-20 dB).In addition,we found that the performance of ZF precoding at 6 GHz is closer to the theoretical derivation.Therefore,when the Massive MIMO channel does not reach a favorable propagation state,the ZF precoding performance can be improved by increasing the value of SNR or the antenna array spacing,or using a more suitable carrier fre-quency.In summary,we conducts experimental measurements and theoretical anal-yses around the advantageous transmission characteristics of Massive MIMO channels.In detail,the Massive MIMO channel model,propagation character-istics,applicable scenarios,spatial structure,precoding performance,and the advantages over the traditional MIMO are studied and evaluated from the as-pects of statistic results of propagation parameters,channel capacity and eigen-value spread.The research work and results in this article provide a promising direction and contribute to a wider application of Massive MIMO in 5G NR.