3D MIMO Channel Propagation Characterics And Modeling

The rapid development of mobile smart devices and application of new ser-vices in fifth-generation(5G)communication require a higher transmission data rate,lower latency and higher power consumption efficiency.3 Dimensional MIMO(3D MIMO),which expands the traditional antenna array from hori-zontal domain to three-dimensional space,can achieve a higher system spectral efficiency and has been regarded as a key technology.Different to 4G which adopts 2D channel model,in order to better evaluate 3D MIMO in 5G,we car-ried out external field measurements under multiple scenarios and make analy-sis in measurements and theory.Specifically,innovations are as follows:1.Based on 3D MIMO channel measurement and data post-processing,we present statistical results of 3D channel propagation.It's essential to perform measurements to acquire accurate knowledge of channel trans-mission characteristics.Using the MIMO channel measurement platform(Sounder)and 3D antenna arrays,we conducted extensive field measure-ments at 3.5 and 6 GHz under different scenarios,such as typical indoor meeting room,indoor work cubic,urban micro-cellular outdoor-to-indoor(021)and urban macro-cellular 021,etc.In measurements,an orthog-onal polarized uniform planar array with 32 antenna elements was used as transmitter,a 56-pole orthogonal polarized onmidirectional cylindrical array was used as receiver.Through data post-processing,we obtain 3D channel impulse response(CIR)and extract channel propagation param-eters,including multipath power,delay,horizontal angles,vertical angles and channel polarization components.Furthermore,we present their sta-tistical characteristics such as the mean and variance,and fit the parameter distributions.Results show that elevation angle obeys Laplacian distribu-tion,elevation angle spread follows LogNormal distribution,signals at 6 GHz is better scattered in line-of-sight(LoS)short distance propagation while suffer serious pathloss at long distance propagation,which result in much less received multi-paths as comparaed to that of 3.5 GHz.2.Validation and Analysis of 3D MIMO based on measurements.It still remains unknow whether the currently standardized 3D channel model can fit the actual channel well and to what degree it outperforms 2D chan-nel model.Thus,we reconstruct 2D and 3D CIR and validate the perfor-mance of 3D channel in terms of channel capacity and eigenvalue dis-tribution.The research methods are as follows:Firstly,substitute the extracted propagation parameters and practical antenna radiation pattern into the standardized 3D MIMO channel model,thus we can reconstruct 3D CIR.Similarly,reconstructed 2D CIR can be obtained.Thus,using the obtained real CIR from measurements,the reconstructed 2D and 3D CIRs,we calculate the eigenvalues of channel spatial correlation(SC)and capacity.Results demonstrate that 3D channel capacity and eigenvalues fit quite well with corresponding results of real measurements,which indi-cates that 3D channel model simulates the real channel accurately.Com-pared with 2D channel,channel capacity of 3D channel is much higher and eigenvalues distribute more evenly,thus 3D MIMO has a stronger ability of parallel transmission.Furthermore,the influence of different antenna structures and antenna positions on system performance are studied.3.Theory comparison of 2D and 3D MIMO channel models in terms of SC and channel capacity.Although we have validated the superiority of 3D channel over 2D counterpart based on measurements as mentioned above,theory analysis and quantitative results are still lacking.There-fore,to analyze effects of elevation domain on system performance,we has decomposed 3D SC to horizontal and vertical domains,and presented a general method in deriving the closed-form expression of 3D SC via sine function approximation and Bessel series expansion,which solves the problem of double integrations invovled with complicated components.Furthermore,we has derived the relationship between 2D and 3D SC the-oretically,results show that 2D SC is a special case,i.e.,when the antenna vertical interval spacing(VIS)is 0 and the multipath only propagates in the azimuth domain,3D SC degenerates to 2D SC.Meanwhile,the SC gap between 2D and 3D SC grows squarely with the VIS and elevation angle spread(EAS)of multipaths.Finally,the loose and tight bounds of the 3D channel capacity are given.The loose upper bound is independent inden-tical distribution(i.i.d)channel capacity,while the loose lower bound is 2D channel capacity.4.Investigate 3D MIMO multi-user interference and propose an inter-ference mitigation algorithm via 3D space separation.Concerning the line-of-sight(LoS)multi-user interference within a single cell,firstly,we present the system model and derive statistical metrics of the multi-user interference,including the first-order center moment(mean)and second-order center moment(variance),where the interference decreases rapidly with multi-user EAS.Furthermore,for general 3D MIMO channel(both LoS and non-LoS components),we investigate influences of distance from base station to users,the interval spacing among users,and angular spread of multi-user distribution on signal to interference plus noise ratio(SINR).Results demonstrate that cells with small coverage radius can resist the serious path loss,however,dense multi-user in a narrow area can cause serious interference,therefore,we proposes a user scheduling algorithm based on multi-user interval spacing,which divides the users into groups with maximal intra-group distance,thus a much higher system sum rate is achieved by fully utilizing the 3D space.In summary,we have evaluated 3D MIMO in terms of statistic results of 3D channel propagation,channel capacity,eigenvalue distribution,2D vs 3D SC and interference mitigation,etc.,both in measurement observations and the-ory analysis.Collectively,all work in this article will contribute to the promis-ing future of 3D MIMO utilization in 5G.