KR20200051336A - method for channel of massive air-channel simulator - Google Patents

Abstract

The present invention relates to a method of creating a channel of a channel simulator for massive MIMO capable of creating a channel for massive MIMO by expanding a statistical channel model based on an existing antenna correlation in a channel simulator creating a channel between a terminal and a massive MIMO base station including a plurality of antennas. According to a first feature of the present invention, a channel simulator for massive MIMO includes: a body part including a UE I/F board and a BS I/F board performing RF signal processing on a base station including at least ten antennas and a terminal including at least one antenna and an LP board performing fading processing for each channel between the boards; and a scenario server controlling the LP board such that an MIMO channel is created in accordance with a test scenario. In this channel simulator for massive MIMO, the method is performed in accordance with the control of the scenario server, but a massive MIMO channel is created based on a statistical antenna correlation and an LOS component and an NLOS component are all calculated through an antenna array manifold.

Description

How to create a channel for a massive MIMO channel simulator {method for channel of massive air-channel simulator}

The present invention relates to a method for generating a channel in a channel simulator for a massive MIMO, and in particular, a channel model for generating a channel between a massive MIMO base station and a terminal having a plurality of antennas is extended with a statistical channel model based on a conventional antenna correlation. It relates to a channel generation method of the channel simulator for the massive MIMO so that the channel for the massive MIMO can be generated.

In recent years, the amount of wireless data required by the exponential growth of smart devices has also increased rapidly, but since the available frequency bandwidth and wireless channel capacity are limited, efficient utilization of existing radio resources (frequency, etc.) It is demanded stronger than ever.

Massive MIMO (Massive Multiple Input & Multiple Output) is a multi-antenna technology that is equipped with dozens of antennas, more than the current ones, to achieve high energy efficiency along with high transmission speed.

As a result of the first research on massive MIMO in a time division multiplexing (TDD) system, that is, if a perfect channel vector can be obtained by using up / downlink channel correlation, the greater the number of antennas, the more interference between different users' channels. Since the announcement that it is possible to service multiple users at the same time even if a simple transmit / receive filter is used by canceling, research on massive MIMO has been actively conducted.

Accordingly, in an international standardization organization such as 3GPP (3rd Generation Partnership Project), as part of an attempt to improve the performance of the LTE-Advanced system, the concept of beam forming is used to allow multiple users to use the same radio resources simultaneously, and the radio channel capacity of the base station cell. FD-MIMO (Full Dimension (or 3D) MultipleInput & MultipleOutput) technique, which can maximize the sum rate, is adopted as a standard through release 13. Hereinafter, the term 'massive MIMO' is used as a comprehensive concept including FD-MIMO as a subset, and the massive MIMO technology is predicted to be adopted as an essential technology in a 5G system currently under development.

In this massive MIMO system, it was agreed that a channel model and a generation method considering a phase component of a multipath signal are required, and in addition, AOA (angle of incidence, angle of the signal) through a channel model such as 3D-SCM (3-Dimensional Spatial Channel Model) It is recommended to develop and apply a channel model that can reflect characteristics such as of Arrival (AOD) / AOD (Angle of Departure).

On the other hand, since the radio wave environment existing in the radio channel is very diverse, the original performance of the radio system must be properly demonstrated in each different radio wave environment. In order to guarantee the performance of the radio system, the simulation and analysis are verified as well as the protocol. Typing and field testing should be done. However, field testing of the developed wireless system under all environmental conditions takes a lot of time and money, so a real-time channel simulator can be used as a more practical method. This is a system that can simulate almost any environment that can actually occur in a wireless channel.

Accordingly, the applicant has a large capacity configured to easily apply bidirectional path loss and bidirectional real-time fading to all paths (P * Q) between P (> 2 integer) base stations and Q (> 2 integer) terminals. The channel simulator was patented and registered as a patent number 1286023. Since then, related studies have been continued. Recently, a channel simulator for massive MIMO has been developed. Hereinafter, the 'massive MIMO base station' is defined as a base station having 10 or more antennas, and the 'channel simulator for massive MIMO' is defined as a simulator in which one or more massive MIMO base stations can be connected.

On the other hand, since the massive MIMO channel is based on a beamforming concept, not only the size of the signal between the antennas, but also the accuracy and stability of the phase synchronization act as a limitation of performance. Therefore, in the channel simulator for massive MIMO, the channel generation performance that can secure the terminal mobility stably and accurately is very important. In addition, since the base station needs to adaptively track the channel capacity according to the change of time using signals or information received from the terminal, the downlink channel and the uplink channel must accurately reflect the actual air channel.

However, the conventional MIMO system defined in the previous standard such as 3GPP does not utilize the phase information for each antenna of a multipath signal incident and radiated to an antenna of a plurality of terminals or each terminal. There was a problem that can not be generated.

On the other hand, in the case of the currently recommended 3D-SCM technology, the related variables are somewhat more than the channel generation method used in the conventional MIMO system, that is, the statistical correlation based method is somewhat more and the conditions for the occurrence method are In addition to being diverse and complex, there was a problem that the handover function was not supported.

Above all, since 3D-SCM technology uses a completely different technology from the conventional statistical correlation-based channel generation method, there is no existing accumulated data (data) and, accordingly, through the channel generated by 3D-SCM technology. There was a problem that it was not easy to verify the test results.

Prior Art 1: Registered Patent No. 10-1286023 (Invention name: Channel Simulator) Prior Art 2: Registered Patent No. 10-1606354 (Invention name: calibration method of channel simulator) Prior art 3: Patent application for 10-2017-0077671 (Name of invention: Control method of channel simulator)

The present invention has been devised to solve the above-mentioned problems, and extends a statistical channel model based on a conventional antenna correlation in a channel simulator for generating a channel between a massive MIMO base station and a terminal having a plurality of antennas, and massive MIMO. It is an object of the present invention to provide a method for generating a channel of a massive MIMO channel simulator that can generate a channel.

According to a first aspect of the present invention for achieving the above object, BS I / F board and UE I / F that perform RF signal processing for a base station having 10 or more antennas and a terminal having one or more antennas, respectively A main body having a board and an LP board performing fading processing for each channel between the boards; Performed according to the control of the scenario server in the channel simulator for the massive MIMO including the scenario server that controls the LP board so that the MIMO channel according to the test scenario can be generated, but generates the massive MIMO channel based on the statistical antenna correlation. , A method for generating a channel of a massive MIMO channel simulator for calculating both the LOS component and the NLOS component using an antenna array manifold is provided.

According to a second aspect of the present invention, each of the BS I / F board and the UE I / F board and each of these boards performs RF signal processing for a base station having 10 or more antennas and a terminal having one or more antennas, respectively. A main body unit equipped with an LP board for performing fading processing for each channel; Massive MIMO channel simulator including a scenario server that controls the LP board so that the MIMO channel according to the test scenario set through the test manager and the test manager can be set up according to the control of the calibration server so that the user can set the desired test scenario. Performed, desired channel model and location of base station and terminal, number of taps in multi-path, angle of incidence and radiation angle of each multi-path, movement speed of a terminal required to generate a mobile channel, or a Doppler frequency corresponding to the same, type of PAS and its dispersion (A) selecting or setting an antenna correlation between two antennas adjacent to each other; (B) calculating the power ratio K of the LOS component and the NLOS component by the antenna correlation diagram and applying the antenna array manifold from the first multi-path to calculate the channel coefficients for the LOS component and the NLOS component, respectively, and then adding them A method for generating a channel of a massive MIMO channel simulator, including the step (c) of calculating channel coefficients for each antenna for a corresponding multipath, is provided.

In the second feature described above, the channel model includes Ped A / B, Veh A / B, and EPA / EVA / ETU, and the number of taps (N _t ) of the multipath and the corresponding power delay (PDP) through channel model selection Profile).

이고,

는 상기 안테나 상관도이다.

ego,

Is the antenna correlation diagram.

은,Channel coefficient vector of the NLOS component

silver,

에 의해 산출되고,

는

에 의해 산출되며,

는

에 의해 산출되고,

은 상기 PAS의 종류에 의해 선택되며 도플러 주파수만큼 대역 제한된 가우시안 분포를 갖는 시변채널 신호이다.

Is calculated by,

The

Is calculated by,

The

Is calculated by,

Is a time-varying channel signal selected by the type of the PAS and having a Gaussian distribution band-limited by the Doppler frequency.

When the communication method is FDD, a channel in the opposite direction is calculated by substituting a transformation matrix stored in advance after step (c).

According to the channel generation method of the channel simulator for the massive MIMO of the present invention, it is possible to generate a new channel model that can be applied to the massive MIMO channel through a simple modification of the existing channel model. It can maintain compatibility with the model. Accordingly, it is possible to consistently analyze the performance of a new wireless communication system called massive MIMO technology while succeeding to the vast test results obtained by utilizing the existing channel model.

1 is an overall system configuration diagram of a channel simulator for massive MIMO to which the channel generation method of the present invention can be applied.
Figure 2 is a flow chart for explaining the channel generation method of the channel simulator for the massive MIMO of the present invention.
3A and 3B are graphs showing the PAS for the first multi-path including the LOS component and the NLOS component to such an extent that the K value is significantly small.
4A and 4B are graphs showing the PAS for the first multi-path including the LOS component and the NLOS component to a significantly smaller K value according to the channel generation method of the present invention.
FIG. 5A is a PAS graph showing an FDD downlink channel using a transformation matrix and showing the result, and FIG. 5B is a PSS graph showing an enlarged section of FIG. 5A.

Hereinafter, with reference to the accompanying drawings will be described in detail a preferred embodiment of the channel generation method of the channel simulator for the massive MIMO of the present invention.

1 is an overall system configuration diagram of a channel simulator for massive MIMO to which the channel generation method of the present invention can be applied. As shown in FIG. 1, a channel simulator for a massive MIMO to which the channel generation method of the present invention can be applied is largely a massive MIMO base station (eNB; enhanced Node-B or BS; base station) 600 and a user equipment (UE). Link processor board (hereinafter referred to simply as 'LP board') which performs real-time fading processing, that is, slow fading and fast fading processing, for a plurality of channel signals connecting Equipment or MS; Mobile Station (700) 700 (110), a base station interface board (base station interface board; hereinafter simply referred to as 'BS I / F board') connecting the base station 600 and the LP board 110 (120), the LP board 110 and the terminal ( 700) connected to the terminal interface board (User Equipment Interface Board; hereinafter simply referred to as 'UE I / F board') 130, the simulator body portion (hereinafter simply referred to as 'body portion') 100, The test scenario you want, for example The location or distance of the base station 600 and the terminal 700 or the direction of each antenna, the number of multi-paths (scatterers), PAS (Power Angular Spectrum); A test manager (TM) 300 that supports setting the degree of variance of the received power distribution profile according to the azimuth of the shock waves and the movement speed of the terminal based on a graphical user interface (GUI) It may include a scenario server (Scenario server; S-server) 400 for controlling the LP board 110 so that the MIMO channel according to the test scenario set by the user through the manager 300 can be generated.

The embodiment of FIG. 1 shows a channel simulator for 64 * 64 massive MIMO in which one passive MIMO base station 600 having 64 antennas and 16 terminals 700 having four antennas are connected. Here, when the BS I / F board 120 is composed of 16 and the UE I / F board 130 is composed of 16, four BS antennas are connected to each BS I / F board 120 and each UE One terminal 700 may be connected to the I / F board 130. Furthermore, it is also possible to expand the channels that can be created with 128 * 64.

On the other hand, in a multi-path communication channel, a direct wave (LOS: Line Of Sight) component, a non-linear wave (NLOS: Non Line Of Sight) component, a reflected wave component, and a diffraction wave component, etc. are mutually disposed between the transmitter base station and the receiver terminal. It exists with an impact. That is, since the direct wave and non-direct wave components are not only received by the terminal through multiple paths, but also because of the movement of the terminal, Doppler spread occurs, so that the mobile communication is placed in a poor propagation environment compared to the fixed communication.

In general, a rural or suburban environment in which a direct wave signal is present can be described as a Ricean channel model, and a downtown area where the direct wave signal is sparse and has many composite signals by multipath can be described as a Rayleigh model. Can be. In addition, there is a shadowing effect due to non-uniformity of the surrounding terrain.

On the other hand, when generating a massive MIMO channel, it is very important to create a fading channel by a traditional time delay and a channel by spatial scatterer distribution. The channel model of the existing international standard, ITU / 3GPP, is capable of sufficiently evaluating the performance of temporal fading channels, and its technical implementation is also easy. On the other hand, for passive MIMO channels using spatial channel characteristics, PAS maintenance characteristics or FDD channels are used. It did not provide enough channel generation variables to satisfy the channel reciprocity.

Accordingly, in the present invention, a method for generating a channel so that accumulated test results can be inherited by presenting an evolution method from a conventional MIMO standard channel to a massive MIMO channel using a statistical channel model based on antenna correlation Explain in detail.

2 is a flow chart for explaining a method for generating a channel of a massive MIMO channel simulator according to the present invention, and may be performed mainly using a scenario server.

As shown in FIG. 2, first, in step S10, a user selects or sets a desired channel model and a location of a base station and a terminal through a test manager, and the location of the base station and a terminal can be set based on GUI. On the other hand, the channel model selectable in step S10 may be a channel model defined in ITU / 3GPP, for example, Ped A / B, Veh A / B and EPA / EVA / ETU, as in the conventional MIMO channel generation method. , Through this selection, the number of taps (N _t ) of a multi-path and a corresponding Power Delay Profile (PDP) may be determined together.

는 인접한 두 안테나 사이의 상관도를 나타내는 것으로, 안테나의 위치에 관계없이 단일 값으로 설정된다.Next, in step S20, the position of the scatterer, that is, the angle of incidence and the angle of radiation, the moving speed of the terminal required to generate the mobile channel, the Doppler frequency corresponding to the type, the degree of dispersion of the PAS and the degree of dispersion thereof, or can be obtained by inverse Fourier transform Select or set antenna correlation. Here, the antenna correlation diagram

Denotes a correlation between two adjacent antennas, and is set to a single value regardless of the position of the antenna.

 The PASs that can be set in step S20 include, for example, a square wave spectrum, a Jakes spectrum, or a Laplacian spectrum, which are a sync function, a zero order Bessel function, or an exponential ) Function.

는 아래의 수학식 1과 같이 NLOS 및 LOS 성분의 조합으로 모델링하며 나머지 다중 경로는

인 NLOS 성분으로만 모델링된다.Meanwhile, the MIMO channel correlation matrix of the first multipath among multipaths

Is modeled as a combination of NLOS and LOS components as shown in Equation 1 below.

It is modeled only with the phosphorus NLOS component.

의 성분은 PAS를 통해 규정할 수 있는데, 이를 역푸리에 변환하여 얻을 수 있는, 아래의 수학식 2와 같은 안테나 상관 함수로 표현하기도 한다.In Equation 1, NLOS channel correlation matrix (channel matrix)

The component of can be defined through PAS, which is also expressed by an antenna correlation function as shown in Equation 2 below, which can be obtained by inverse Fourier transform.

In the method of the present invention, for simplicity, only the transmit antenna correlation is described. However, even if the receive antenna correlation is considered at the same time, the same analysis can be performed by a simple extension method.

는 직접파이므로 아래의 수학식 3과 같이 랭크 1(rank 1) 행렬로 모델링할 수 있다.Meanwhile, in Equation 1, the LOS channel correlation matrix

Since is a direct wave, it can be modeled as a rank 1 matrix as shown in Equation 3 below.

는 입사 신호에 대한 i번째 안테나 응답인 어레이 매니폴드(array manifold)를 나타내며,

는 어레이 매니폴드 벡터의 에르미트 행렬(Hermitian matrix)인 (

) 곱에 의한 랭크 1 행렬로 주어진다.In Equation 3

Denotes the array manifold, i-th antenna response to the incident signal,

Is the Hermitian matrix of the array manifold vector (

) Given by a rank 1 matrix by product.

가 되고,

, 즉 안테나 상관도가 1이 된다는 것을 알 수 있다. 이때 NLOS 채널 상관 행렬

는 아래의 수학식 4와 같이 표현된다.On the other hand, K, which represents the power ratio of the LOS component to the NLOS component, and the antenna correlation have a close relationship. If only the LOS component is present, that is, if K increases infinitely

Become

That is, it can be seen that the antenna correlation is 1. At this time, the NLOS channel correlation matrix

Is expressed as Equation 4 below.

=1이 되어 안테나 보어 사이트 벡터(antenna bore-sight vector)로 구성되는 모양이 되며, LOS 성분의 입사 방향과 전혀 관계없는 벡터(방향)을 갖게 된다.In other words, rank

It becomes = 1, which is a shape composed of an antenna bore-sight vector, and has a vector (direction) that is completely independent of the incident direction of the LOS component.

 At this time, if K is very large, that is, if the LOS component is very large compared to the NLOS component, there is no problem, but when K is significantly reduced, a corresponding PAS difference occurs between the LOS component and the NLOS component.

3A and 3B are graphs showing the PAS for the first multipath including the LOS component and the NLOS component to such a degree that the K value is significantly small, and are PAS graphs when the incident angle and the reflection angle are both 22.5 degrees. 3A is a PAS graph when the antenna correlation diagram is 0.30, and FIG. 3B is a PAS graph when the antenna correlation diagram is 0.99.

As shown in FIG. 3, although the magnitude and dispersion degree vary according to the antenna correlation, in the case of the LOS component (H _rc ), the central angle of the PAS is formed at exactly 22.5 degrees, whereas the LOS component and It can be seen that the PAS of the NLOS component (H _rl ), which should be formed to have a similar center angle, does not sufficiently provide the desired incidence / radiation direction information because the center angle is always toward 0 degrees, which is the boresight.

In order to solve this problem, in the present invention, Equation 1 is modified as shown in Equation 5 below so that it can be applied to all multipath channels regardless of whether LOS components are present.

Equation (5) may obtain simplified results as in the case of Equations 6 to 8 as in the existing channel model.

와 K를 통합적으로 해석할 수 있도록 지원하는데, 이에 따라 실제적으로는 불가능한 필드 환경에 대한 채널 생성을 미연에 방지할 수 있다.According to the channel generation method of the present invention, the equations 6 and 8 provide the same results as before, and in the case of the equation 7, that is, an intermediate value in which an antenna correlation is not 0 or 1 in an environment in which an NLOS component is present Even if exists as, the channel can be generated such that the incident angle / radiation angle set by the user is the central angle of the PAS graph. In particular, rather than a clear distinction between the NLOS component and the LOS component, the correlation between two adjacent antennas by Equation 9 below

It supports the integrated analysis of and K, and accordingly, it is possible to prevent the creation of a channel for a field environment which is practically impossible.

에서 K=

가 되고,

에서 K=0이 되며,

의 범위에서는 함수 관계가 성립함으로써 안테나 상관도의 크기만으로도 그에 맞는 다양한 채널 생성이 가능해진다.In Equation 9

K =

Become

In K = 0,

In the range of, since a functional relationship is established, it is possible to generate various channels according to the size of the antenna correlation.

Returning to FIG. 2 again, in step S30, K is calculated by substituting the set antenna correlation into Equation (9), and a path index (i) for repeatedly performing channel generation for each multipath to sequentially generate channels from the first multipath _TAP ) Set to 0. In step S40, it is determined whether the path index (i _TAP ) is 0. If it is 0, since it corresponds to the first multipath, in step S50, the LOS component included in the first multipath is calculated using the antenna array manifold.

은 아래의 수학식 10과 같이 입사/방사각에 의해 계산되는 어레이 매니폴드 채널 벡터

에 도플러 주파수 성분

가 포함된 시간 채널

를 곱하여 생성한다.

는 초기 랜덤 위상 값이며

는 채널 샘플링 주파수이다.Specifically, LOS channel coefficient vector

Is an array manifold channel vector calculated by the incident / radiation angle as in Equation 10 below.

Doppler frequency component

Time channel containing

Multiply by to produce.

Is the initial random phase value

Is the channel sampling frequency.

은 전통적인 채널 계수 생성 방법을 따른다. 이는 아래의 수학식 11과 같이 상관 행렬

을 갖는 랜덤 변수를 상관 행렬

(항등 행렬)를 갖는 랜덤 벡터

에

를 곱해서 만드는 방법이다.

은 앞서 열거한 PAS 스펙트럼의 모양에서 선택되며 도플러 주파수만큼 대역 제한된 가우시안(Gaussian) 분포를 갖는 시변(time-varying) 채널 신호다. 또한, 개별 다중 경로마다 신호의 입사각/방사각 정보를 적용하기 위해 대각 행렬

를 적용한다.When the channel coefficients for the LOS component are calculated in this way, in step S70, the channel coefficients for the NLOS component of the corresponding multipath are calculated using the corresponding antenna array manifold.

Follows the traditional channel coefficient generation method. This is a correlation matrix as shown in Equation 11 below.

Correlation matrix of random variables with

Random vector with (identity matrix)

on

This is how to multiply by.

Is a time-varying channel signal that is selected from the shapes of the PAS spectrum listed above and has a Gaussian distribution band-limited by the Doppler frequency. In addition, a diagonal matrix is applied to apply the incident angle / radiation angle information of each signal for each multipath.

Apply.

In step S80, as shown in Equation 12 below, the LOS channel vector and the NLOS channel vector are summed to calculate channel coefficients per antenna for the corresponding multi-path, and the channel coefficients of each multi-path must be applied to the signal independently.

In this way, after the channel calculation for each antenna for the first path is completed, step S90 is performed to increase the path index (i _TAP ) by 1, and in step S100, the path index (i _TAP ) is the number of multipath taps set by the user ( N _t ). As a result of the determination in step S100, if the path index i _TAP is less than or equal to the number of multi-path taps N _t , the process returns to step S40 to determine whether the path index i _TAP is not 0, that is, not the first path.

As a result of the determination in step S40, when the first path is not the only NLOS component exists, the process proceeds to step S60 to process the K value as 0 and the LOS component also processes to 0, and then proceeds to step S70 to perform the corresponding antenna array. The manifold is used to calculate the NLOS component of the multipath.

As a result of the determination in step S100, when the path index i _TAP is greater than the number of multipath taps N _t set by the user, the process proceeds to step S120 to determine whether the communication method is FDD.

In this regard, in the case of the FDD method, in a channel model such as 3D SCM proposed by 3GPP, the downlink channel and the uplink channel have the same spatial (channel spatial reciprocity) condition based on the measurement of the actual channel. ) That is, the channel model that has the same incident angle / radiation angle and only the center frequency of the signal fluctuates is applied. In this case, each multipath has a different angle of incidence / radiation, but the CR model of the same method is applied.

In view of this, the same method can be applied to a traditional statistical channel model. In the present invention, an arbitrary one-way channel, for example, uplink, can be easily obtained through an array manifold transformation matrix calculated when downlink and uplink frequencies are determined. A method of determining a downlink channel based on a channel is proposed.

는 입사각/방사각(θ)과 신호 주파수(

)의 함수인데, FDD 방식의 경우 업링크 주파수와 다운링크 주파수가 다르기 때문에 업링크 채널을 통해 추정한 값을 다운링크 채널에 그대로 적용할 수 없다.As shown in Equation 3 above, the array manifold of the antenna

Is the incident angle / radiation angle ( θ ) and signal frequency (

). In the case of the FDD method, since the uplink frequency and the downlink frequency are different, the estimated value through the uplink channel cannot be applied to the downlink channel as it is.

에 대해 다운링크 어레이 매니폴드

를 생성하기 위해 아래의 수학식 13과 같은 변환 행렬 T를 고려한다.Accordingly, in the present invention, the uplink array manifold at a given frequency

About Downlink Array Manifold

Consider the transformation matrix T as in Equation 13 below to generate.

In this case, if θ is applied and merged for all multipath incident / radiation angles, it can be expressed as Equation 14 below.

In Equation 14, T is a matrix for converting downlink channels and uplink channels having the same incident angle / radiation angle but different frequencies. If Equation 15 is replaced with a value satisfying the LS (Least Square) condition, Equation 15 below Is saved.

은 의사 역행렬(pseudo inverse matrix)인데,

(P는 안테나 개수)일 경우에도 그 존재를 보장하기 위해

와 같이 대각 부하(diagonal loading)를 적용할 수 있다.In Equation 14

Is a pseudo inverse matrix,

(P is the number of antennas)

As shown, diagonal loading can be applied.

When this transformation matrix is applied, the downlink that satisfies the CR condition is multiplied by the matrix using the transformation matrix for the uplink channel stored in advance without calculating and storing the new array manifold individually for each uplink and downlink or multipath. Link channels can be quickly calculated in real time.

Returning to FIG. 2 again, as a result of the determination in step S110, when the communication method is not FDD and TDD, since the uplink and downlink carrier frequencies are the same, the program is immediately terminated, whereas in the case of the FDD method, the process proceeds to step S120. As described above, the pre-stored transformation matrix is substituted to calculate the channel in the opposite direction, and then the program ends.

4A and 4B are graphs showing the PAS for the first multi-path including the LOS component and the NLOS component to a significantly smaller K value according to the channel generation method of the present invention, for a case where both the incident and reflection angles are 22.5 degrees It is a PAS graph. 4A is a PAS graph when the antenna correlation diagram is 0.30, and FIG. 4B is a PAS graph when the antenna correlation diagram is 0.99.

As shown in FIG. 4, according to the channel generation method of the present invention, by applying an array manifold diagonal matrix to the NLOS component as shown in Equation 5 above, the center angle (incidence / reflection angle) of the PAS for both the LOS component and the NLOS component It can be seen that it is formed at 22.5 degrees, and as a result, it is possible to efficiently form a channel including an NLOS component having a desired incidence / radiation angle.

FIG. 5A is a PAS graph showing the result of generating an FDD downlink channel using a transformation matrix, and FIG. 5B is a PSS graph showing an enlarged section of FIG. 5A with downlink and uplink frequencies of 100 MHz. If present, it indicates the change in PAS for the downlink channel.

As shown in Fig. 5, for example, when the angle of incidence and the angle of incidence are 14.2 degrees and 37.1 degrees, respectively, the correct calculation values for each frequency of the array manifold (see the green graph) and the calculation values by the transformation matrix (see the red graph) It can be seen that the PAS of is consistent within the error range.

Or more, with reference to the accompanying drawings, a preferred embodiment of a method for generating a channel of a massive MIMO channel simulator of the present invention has been described in detail, but this is only an example, and various modifications and changes can be made within the scope of the technical idea of the present invention. will be. Therefore, the scope of the present invention should be defined by the following claims. For example, in the above-described embodiment, the description has been made using LTE as an example, but may be applied to future 5G technologies within the scope of maintaining the technical idea.

In addition, terms such as 'board' or 'kit' are only arbitrarily borrowed for convenience of logical or functional explanation, and should not be used for limiting the scope of rights. It may be described separately in units.

100: simulator body, 110: link processor (LP) board,
120: base station interface (BS I / F) board,
130: terminal interface (BS I / F) board,
300: test manager, 400: scenario server,
600: Massive MIMO base station, 700: Terminal

Claims (6)

BS I / F board and UE I / F board that perform RF signal processing for a base station having 10 or more antennas and a terminal having 1 or more antennas, respectively, and an LP board that performs fading processing for each channel between these boards Body portion provided with; Performed according to the control of the scenario server in a massive MIMO channel simulator including a scenario server that controls the LP board so that a MIMO channel according to the test scenario can be generated.
A method of generating a channel in a channel simulator for a massive MIMO, wherein a massive MIMO channel is generated based on a statistical antenna correlation, and both LOS components and NLOS components are calculated using an antenna array manifold. BS I / F board and UE I / F board that perform RF signal processing for a base station having 10 or more antennas and a terminal having 1 or more antennas, respectively, and an LP board that performs fading processing for each channel between these boards Body portion provided with; Massive MIMO channel simulator including a scenario server that controls the LP board so that a MIMO channel according to the test scenario set through the test manager and the test manager can be set according to the control of the calibration server so that the user can set the desired test scenario. Performed,
Desired channel model, location of base station and terminal, number of taps of multi-path, angle of incidence and radiation angle of each multi-path, moving speed of a terminal required to generate a mobile channel, a corresponding Doppler frequency, type of PAS, and degree of dispersion or adjacent (A) selecting or setting an antenna correlation between two antennas;
(B) calculating a power ratio K of the LOS component and the NLOS component by the antenna correlation diagram, and
After applying the antenna array manifold from the first multi-path, the channel coefficients for the LOS component and the NLOS component are respectively calculated, and then summed to calculate the channel coefficients for each antenna for the corresponding multipath (c). How to create a channel in the channel simulator. The method according to claim 2,
The channel model includes Ped A / B, Veh A / B and EPA / EVA / ETU, and the number of taps (N _t ) of the multi-path and the corresponding Power Delay Profile (PDP) are determined through channel model selection. Channel generation method of the channel simulator for massive MIMO, characterized in that. The method according to claim 3,

ego,

A channel generation method of a channel simulator for massive MIMO, characterized in that the antenna correlation diagram. The method according to claim 4,
Channel coefficient vector of the NLOS component

silver,

Is calculated by,

The

Is calculated by,

The

Is calculated by,

Is a time-varying channel signal selected by the type of the PAS and having a Gaussian distribution band-limited by a Doppler frequency. The method according to claim 5,
When the communication method is FDD, the channel generating method of the massive MIMO channel simulator is calculated by substituting a transformation matrix stored in advance after step (c) and calculating a channel in the opposite direction.

Images