Font Size: a A A

Research On 5G Millimeter-wave Channel Propagation Characteristics And Modeling

Posted on:2022-05-29Degree:DoctorType:Dissertation
Country:ChinaCandidate:T JiangFull Text:PDF
GTID:1488306326979459Subject:Information and Communication Engineering
Abstract/Summary:PDF Full Text Request
With the vigorous development of intelligent services,high-definition video streaming,cloud services,and other services,mobile communication net-works are facing huge challenges in terms of key performance such as through-put,delay,and peak rate.In response to these challenges,major standardiza-tion organizations,research institutions,and commercial companies around the world have researched the Fifth Generation Wireless Communication System(5G).Millimeter-wave communication is a key technology of 5G systems.Its abundant spectrum resources and continuous bandwidth of several gigahertz have driven a substantial increase in the capacity of communication systems.However,due to the increase of the carrier frequency,the delay domain char-acteristics,spatial angle characteristics,and loss characteristics of the wireless channel which are frequency-dependent will change greatly.The channel char-acteristics determine the optimum performance of the wireless system.To obtain the best system performance,the wireless channel must be fully studied.In this paper,we study the delay characteristics,spatial angle characteristics,path loss,and blockage loss characteristics of 3.5,6.2,and 28 GHz based on the actual measurement results of the indoor corridor,indoor factory,and outdoor Urban Microcell(UMi)scenes.The main content and innovations of this thesis are summarized as follows:(1)Millimeter-wave platform construction and the design of experi-ments.To study the delay characteristics,spatial angle characteristics,path loss,and blockage loss characteristics,this thesis constructs two sets of time-domain channel sounding platforms for measurements.Their working fre-quency bands are 28 GHz and below 40 GHz.Their maximum measurement bandwidths are 800 MHz and 2 GHz,respectively.As for the design of experi-ments,we carried out fixed-point measurements at 3.5 and 28 GHz in corridor and UMi when studying the delay characteristics.By using the ray-tracing method,we delete the effect of different antenna types on the measurement re-sults.In the study of the spatial angle characteristics,we exchanged the receiver and transmitter positions to measure the arrival angles in uplink and downlink.And then we obtained the departure angle and the arrival angle through the channel reciprocity,by which method we can overcome the problem that the millimeter-wave measurements cannot obtain the departure angle and the arrival angle at the same time.When studying the path loss property,we used two sets of antenna heights to analyze the antenna height impact on the path loss model.When studying the characteristics of occlusion loss,we carried out a long-term measurement at 6.2 and 28 GHz in an indoor corridor.The number of people in the corridor was collected in real-time through a camera.Besides,to solve the problem that channel impulse response(CIR)cannot be obtained by dividing the pseudo-noise(PN)signal in the frequency domain,we re-derived the CIR extraction calculation method by using the correlation of the PN sequence.(2)The delay characteristics in millimeter-wave bands.To study the influence of bandwidth on the delay characteristics of millimeter waves,we used a Hanning window to divide the received signal into four different bandwidths from 100 to 400 MHz for analysis.The obtained power delay profile(PDP)is modeled by the Saleh-Valenzuela(SV)model,and the arrival rate of the cluster on the delay,the power attenuation factors of the cluster,and the multipath are obtained.To solve the clustering problem in the SV model,we applied a power reverse search idea and proposed a "bubbling" clustering algorithm.This method solves the delay discontinuity problem in the process of clustering PDP.By comparing the measurement results of 3.5 and 28 GHz,it can be found that when the bandwidth is the same,the multipath power fading speed of millimeter waves is greater than that of centimeter waves in indoor and outdoor scenes.But the cluster power fading speed of millimeter waves is smaller than that of the centimeter wave in the indoor scene.Finally,based on the SV model results,we derived a calculation method of intra-cluster delay spread in continuous form.(3)The spatial angular characteristics in millimeter-wave bands.To study the propagation mechanism of millimeter waves,we analyzed the power angular spectrum(PAS)of arrival angle in the uplink and the downlink.From the results,we can know the direct radiation and first-order reflection are the dominant components in the line-of-sight(LOS)situation in the UMi scene at 28 GHz.The first-order reflection and a small number of second-order reflections are the main propagation under Non-line-of-sight(NLOS).By using the spatial lobe(SL)method we clustered the angle of arrival and obtained the angle spread within the cluster.To verify the different angle generation models in the channel standards,we fitted the clustered arrival angles with the combined PAS and the angle distribution,respectively.It is found that the millimeter-wave horizontal arrival angles are more like Gaussian distribution.Finally,to solve the problem that the angle generation model in the cluster is not flexible enough in the channel standard,a two-dimensional Gaussian model of the angle and power within the cluster was established,and the correlation between the multipath angle and the power in the cluster was analyzed.(4)The path loss and blockage loss characteristics in millimeter-wave bands.To study the path loss model in the Industrial Internet of Things(IIoT),we conducted two measurements in two indoor factories at 28 GHz.By comparing the fitting results of the float-intercept(FI)model and close-in model(CI),we find the FI model is more suitable for the IIoT scenario.To study the crowd blockage loss at 6.2 and 28 GHz,we statistically analyzed the blockage loss of different clusters under various numbers of people.When the number of people is 0,the delays of the clusters at 6.2 and 28 GHz are the same,but the cluster power fading rate of 6.2 GHz is higher than that of 28 GHz.When the number of people in the scene is 1,we used the double-sided edge diffraction model and the simplified four-sided edge diffraction model to simulate the blockage of the whole scene.Through comparison,we can know that the blockage loss of 6.2 GHz is closer to the simplified four-sided edge diffraction model,while the blockage loss of 28 GHz is closer to the double-side diffraction model.Besides,the blockage loss of the whole scene can be divided into two parts:inside and outside the first Fresnel zone.The power distribution outside the first Fresnel zone is mainly concentrated around 0 dB,which is less affected by the frequency.While the power distribution inside the first Fresnel zone is greatly affected by frequency.Since the radius of the first Fresnel zone decreases as the frequency increases,the area of the high loss area at 28 GHz is smaller than 6.2 GHz,but the loss extreme value is greater than 6.2 GHz.When the number of people is greater than 2,the blockage loss will increase with the people number and the propagation distance of the multipath.Finally,by linearly fitting the number of people with the total received power,we can find that the centimeter wave is more sensitive to the number of people in the scene than the millimeter-wave.
Keywords/Search Tags:Millimeter-wave 5G, Channel propagation characteristics, IIoT, Body blockage, SV channel model, Spatial angular model
PDF Full Text Request
Related items