WO2016065683A1

WO2016065683A1 - Three-dimensional beam forming design method in multi-user 3d-multiple input multiple output (mimo) system

Info

Publication number: WO2016065683A1
Application number: PCT/CN2014/092049
Authority: WO
Inventors: 任品毅; 张逸炎; 孙黎
Original assignee: 西安交通大学
Priority date: 2014-10-27
Filing date: 2014-11-24
Publication date: 2016-05-06
Also published as: CN104320169A; CN104320169B

Abstract

Disclosed is a three-dimensional beam forming design method in a multi-user 3D-Multiple Input Multiple Output (MIMO) system. The method includes the following steps: 1) according to a distribution model of users in a 3D scene, calculating a probability density function of vertex angles from a base station to user direct vision diameters, and designing a multi-user scheduling algorithm in a multi-user MIMO system according to the probability density function of the vertex angles; 2) when the base station sends signals to the users, each antenna port in a configured planar antenna array thereof serving an individual user, pre-codes applied to all the antenna ports being pre-codes suggested in a 3rd Generation Partnership Project (3GPP) standard, and the served users being selected from all the users via the multi-user scheduling algorithm; and 3) deciding, by the designed multi-user scheduling algorithm, configurations of the planar antenna array of the base station, and deciding, by the adopted configurations of the base station antenna array, an electronic downward inclination angle used in each pre-code applied to each antenna port, thereby achieving the aim of maximizing the throughput of the system finally. The present invention greatly improves the overall performance of the 3D-MIMO system.

Description

Three-dimensional beamforming design method in multi-user 3D-MIMO system

Technical field

The invention belongs to the field of wireless communications, and in particular relates to a three-dimensional beamforming design method in a multi-user 3D-MIMO system.

Background technique

In recent years, the amount of data required for smart mobile devices is far superior, and 3GPP (3rd Generation Partnership Project) is committed to researching cutting-edge technologies in its standards to improve spectrum efficiency and user experience. Three dimensional Multiple-Input Multiple-Output (3D-MIMO) is one of the key technologies currently studied by 3GPP in the next generation Long Term Evolution (LTE) wireless communication system. Compared with other technologies, 3D-MIMO does not require large-capacity backhaul links and more spectrum resources. It uses a large number of antennas to arrange spatial antenna arrays for three-dimensional beamforming signal transmission, which greatly improves system performance. . However, most of the previous studies have studied three-dimensional beamforming in the traditional two-dimensional channel model. The antenna model used is only a fixed antenna port in which the antenna pattern has been fixed, and is not a planar antenna array composed of an active antenna system in practice. Ignoring the three-dimensional distribution characteristics of the user in space, the method of improving the system throughput only by adjusting the base station antenna downtilt is not ideal in the three-dimensional scene. Therefore, it is necessary to design an efficient three-dimensional beamforming scheme in combination with the characteristics of a complete three-dimensional system suitable for a 3D-MIMO system, particularly the characteristics of a planar antenna array used therein.

Summary of the invention

The object of the present invention is to apply a planar antenna array based on the newly proposed 3D channel model of the 3GPP organization, and propose a three-dimensional beamforming design method for a multi-user 3D-MIMO system for a single-cell multi-user cellular system. The method effectively reduces the interference between multiple users of the system and greatly improves the overall throughput of the system.

To achieve the above object, the present invention adopts the following technical solutions:

A three-dimensional beamforming design method in a multi-user 3D-MIMO system includes the following steps:

1) calculating a probability density function of the apex angle of the base station to the user's direct viewing path according to the distribution model of the user in the 3D scene, And design a multi-user scheduling algorithm in multi-user MIMO system according to the probability density function of the top angle;

2) When the base station sends a signal to the user, each antenna port in the configured planar antenna array serves a separate user. The precoding applied on these antenna ports is the precoding recommended in the 3GPP standard, and the user is served at the same time. Select from all users by a multi-user scheduling algorithm;

3) The configuration of the base station antenna array is determined by the designed multi-user scheduling algorithm. The electronic downtilt angle used in the precoding applied to each antenna port is determined by the configuration of the adopted base station antenna array, and finally the system is maximized. The purpose of throughput.

A further improvement of the present invention is that, in step 1), the application is proposed to distribute the user's distribution law in the new 3D scene in the 3GPP standard, and calculate the probability density function of the apex angle of the user's direct viewing path.

A further improvement of the present invention is that, in step 1), the multi-user scheduling algorithm in the system should satisfy the following conditions:

The difference between the apex angles of the users who are simultaneously scheduled at each time is θ _d , and the expectation that θ _d should be satisfied is as large as possible, that is,

MaxΕθ _d (1)

Where E() indicates that the mathematical expectation is taken, and at the same time, all the θ _d should be made as close as possible, that is,

among them,

Represents the variance of all θ _d .

A further improvement of the present invention is that, in step 2), the following steps are specifically included:

2-1) The base station selects the simultaneously scheduled users at each time interval using the designed multi-user scheduling algorithm;

2-2) The base station uses a rectangular antenna array consisting of M×N antenna elements, one column of each antenna array serving as a single antenna port, serving one of the users simultaneously scheduled;

2-3) The precoding ω _m used by the mth antenna element in each antenna port is

Where M is the number of rows of the antenna array, λ is the carrier wavelength, m is the serial number of the antenna element, d _v is the line spacing of the array, and θ _etilt is the electronic downtilt angle used.

π is the pi, the electronic downtilt angle is the same in the precoding used by all antenna elements in the same antenna port, and the electronic downtilt angle in the precoding used by different ports is different.

A further improvement of the present invention is that, in step 3), the following steps are specifically included:

3-1) The antenna array used by the base station is a rectangular antenna array composed of M×N antenna elements, and the M in the antenna array configuration should satisfy:

among them,

Indicates rounding down, θ _M is the only solution that satisfies the following transcendental equation:

among them,

Ln() means taking the natural logarithm, E() means the mathematical expectation, and θ _d is the difference of the apex angle of the user who is scheduled at the same time;

3-2) The electronic downtilt angle θ _etilt in the precoding used by all antenna elements in the same antenna port is equal and should satisfy

θ _etilt =θ _z +Δθ (5)

Where θ _{z is} the apex angle of the antenna port serving user, and Δθ is the adjustment angle, which should be satisfied

Δθ=θ _M -θ _d (6)

Where θ _d is the difference between the apex angles of the users scheduled at the same time, and θ _M is calculated by the following formula:

Where λ is the carrier wavelength, d _v is the row spacing of the array, and M is the number of rows of the antenna array.

Compared with the prior art, the present invention has the following advantages:

The present invention uses a newly proposed complete three-dimensional system model conforming to the 3D-MIMO system, including a three-dimensional scene, a planar antenna array model and a three-dimensional channel model, and combined with the characteristics of the system model to design a more realistic set of three-dimensional beam assignments. Shape scheme, the influence of user distribution and user scheduling algorithm on beamforming effect is analyzed in the present invention, and The best method, which greatly improves the overall performance of the 3D-MIMO system.

DRAWINGS

Figure 1 is a diagram of a planar antenna array model.

2 is a schematic diagram of beamforming of a single antenna port.

Figure 3 is a system model diagram of the beamforming scheme.

Figure 4 is a comparison of the wideband signal-to-noise ratio of the system under different electronic downtilt angles.

Figure 5 is a comparison of user rates for different antenna models and configurations.

detailed description

The invention will be further described below in conjunction with the drawings and embodiments.

FIG. 1 shows a planar antenna array used by a base station, and the specific description is as follows:

The two-dimensional antenna plane array used by the base station is composed of antenna elements arranged in M rows and N columns, and the distance between rows and columns is d _H and d _{V , respectively} . Each antenna element is independently excited by a transmitter, and its amplitude and phase can be independently adjusted. In LTE, radio resources are allocated in units of antenna ports. Each antenna port is composed of several physical antenna elements, all of which carry the same information. Unlike traditional WINNER channels, in the new 3D system model protocol, the antenna elements, rather than the antenna port, are given by the protocol. The pattern of the antenna elements is divided into horizontal and vertical directions, and they are all modeled as a quadratic function. In the protocol, the direction of the vertical direction of a single antenna element is set to:

θ _3dB =65°, SLA _V =30

Where θ _3dB represents the 3dB beamwidth in the vertical direction, and SLA _V represents the maximum attenuation of the antenna element in the vertical direction.

The protocol states that in an antenna array, every K antennas will form an antenna port, each of which will be weighted by a precoding ω _m . Usually, K = 1 or K = M is specified. In the latter case, a column of M antenna elements in the antenna array will be formed into one antenna port, and the entire antenna array has N ports.

Figure 2 shows a case where an antenna port transmits a signal to a user after precoding ω _m weighting. The precoding recommended in the protocol is used at this time:

Where M is the number of rows of the antenna array, λ is the carrier wavelength, m is the serial number of the antenna element, and d _v is the row spacing of the array.

π is the pi, and θ _etilt is the electron _downtilt defined between 0° and 180°, where 90° is expressed in the plane of the antenna array. The precoding ω _m used by the antenna elements in the same antenna port has the same electronic downtilt angle θ _etilt . Such precoding makes the phase of each antenna element gradually change. We set the phase of the mth antenna element to be offset (exp(-j(m-1)αd _V ), so the gain effect of the entire antenna port can be expressed. F(θ)

Where θ is the pitch angle between the path of the wave propagation and the horizontal plane. When an antenna port transmits information directly to a user, the user's pitch angle θ _e is equal to θ. Therefore, by the summation series formula, F(θ) can be calculated as

The precoding ω _m used can be compared with the assumed phase difference exp(-j(m-1)αd _V )

When expression

When it is established, the directional gain of the port is the largest.

Figure 3 is a model diagram of the entire system. This is a single-cell versatile downlink MIMO system with a base station equipped with a two-dimensional antenna array and Num _u users, all equipped with an omnidirectional antenna. We set x _i and y _i as the relative distances of the i-th user to the base station in the system on the x-axis and the y-axis of the horizontal plane, respectively, and set the heights of the base station and the user i to h ₀ and h _{i , respectively} . Let the user's apex angle θ _z be the angle of the z-axis positive semi-axial base station to the user's direct vision path in the figure, which ranges from 0° to 180°, and according to the geometric relationship

The apex angle θ _z can be calculated from the geographic location of the user, ie

The three-dimensional distribution of the user in the new scene is also proposed in the protocol. In the previous model, all users had a height of 1.5 m. In this case, it is difficult to perform three-dimensional beamforming because the users are too close together in the pitch angle domain to effectively distinguish the beams. In the new three scenarios proposed, the user is still evenly distributed in the horizontal plane as before, but now only 20% of the users are in outdoor users, and the height is 1.5m. Other indoors are considered to be in high-rise buildings. The maximum number of floors in each building is evenly distributed within a certain range, and indoor users are evenly distributed in different heights of the building. Thus, the user's location is limited to a certain area, including the minimum distance from the base station, the maximum distance from the base station, the minimum height of the user, and the maximum height of the user. Thus, from the set user distribution model, the distribution of the user's apex angle θ _z can be calculated, and the probability density function f _Θ (θ _z ) can be obtained.

In the multi-user system, the base station uses one antenna port to serve one user and simultaneously schedules multiple users. Generally, the antenna array uses K=M, that is, each column is one port, and a total of N ports serve N users. In order to maximize the received signal to interference and noise ratio (SINR) of the user after the base station transmits a signal to the user, the received power of the useful signal should be amplified as much as possible. The characteristics of the antenna port pattern composed of antenna elements can be known as the precoded power used by this antenna port. The subtilt angle and the apex angle of the user serving this antenna should have the following relationship: the power to receive the useful signal is maximum:

At this time, the main lobe of the antenna port pattern will be aligned to the service user, and the useful signal gain is the largest, which is obviously optimal when the base station only serves a single user. However, there are more than one user serving at the same time, and the signals between different users will cause interference between users when serving multiple users. In the case of the maximum received power, it is not always guaranteed that the signal to interference and noise ratio is also the largest. Therefore, it is necessary to minimize the power of the interference while maintaining a large signal power, so that the user SINR can be maximized.

According to the direction of the antenna port, we can find that the main lobe is very similar to the quadratic function. Therefore, using the characteristics of the quadratic function - in the middle region is gentle and steep on both sides, we can reduce the interference by adjusting the downtilt angle of the antenna port. When the difference between the top angles of the two users is very small, our strategy of using θ _etilt = θ _z will cause serious interference. At this time, the electronic downtilt angle is shifted by a small angle on the original basis, called the adjustment angle. Δθ, which is characterized by the antenna port pattern, shows that this will cause the signal power received by the user to be slightly reduced and the received interference signal to be greatly attenuated. Using this approach, we can very easily adjust the user's downtilt to get a very large signal to noise ratio gain. The program is expressed by a formula

θ _etilt =θ _z ±Δθ

Among them, the sign indicates the adjustment of the original apex angle, and the direction of adjustment is the direction away from the inclination angle of other users. The electronic downtilt angle will have an optimal value to maximize the user's signal to interference and noise ratio according to the location and top angle of the service user.

At the same time, in the scheme, the configuration M of the base station antenna has a great influence on the performance of the three-dimensional beamforming. Therefore, the value of M will have an optimal value according to the distribution characteristics of the user in the scene.

The user scheduling algorithm has a great influence on the user. In the scheme, the optimal user scheduling algorithm is determined by the distribution of the user in the scene, and the optimal user scheduling algorithm conforms to a certain objective function. The optimal base station configuration and optimal electronic downtilt are related to the optimal user scheduling algorithm.

Figure 4 is a simulation diagram of the cumulative distribution function of the wideband signal to interference and noise ratio of the system under different downtilt schemes. If the shot S is a set of scheduled users then the broadband SINR of the user i can be expressed

Where j represents the interfering user and P _n represents the receiving noise. Broadband SINR is an important indicator of system performance.

Using Monte Carlo simulation, multiple users were selected to use different downtilt angles for transmission in the new scene 3D-Uma defined in the protocol, and the cumulative distribution function curve comparison of broadband SINR under different conditions was obtained. The results show that if we use different constant adjustment angles, the performance of the system will vary greatly. However, no matter how constant the adjustment angle Δθ is comparable to the optimal electronic downtilt angle (optimal adjustment angle) we use, the system performance is maximized under the optimal electronic downtilt angle.

Figure 5 is a graph comparing the cumulative probability density of system user averages and rates for different antenna configurations. When the antenna of the base station uses different configurations, it actually has a greater impact on the performance of the system. It can be seen that under the simulation conditions in the figure, the antenna configuration M achieves the optimal system performance at a certain value, and this value is the optimal M value. At the same time, it can be seen that the system performance improvement is very large compared with the conventionally used antenna port (AP). At the same time, it also shows that the scheme is determined by multiple factors. The global optimal solution of the scheme will be to use the optimal antenna configuration under the optimal user scheduling algorithm, while using the most advantageous electronic downtilt.

Claims

A three-dimensional beamforming design method in a multi-user 3D-MIMO system, comprising the steps of:

1) Calculating a probability density function of the apex angle of the base station to the user's direct vision path according to the distribution model of the user in the 3D scene, and designing a multi-user scheduling algorithm in the multi-user MIMO system according to the probability density function of the apex angle;

2) When the base station sends a signal to the user, each antenna port in the configured planar antenna array serves a separate user. The precoding applied on these antenna ports is the precoding recommended in the 3GPP standard, and the user is served at the same time. Select from all users by a multi-user scheduling algorithm;

3) The configuration of the base station antenna array is determined by the designed multi-user scheduling algorithm. The electronic downtilt angle used in the precoding applied to each antenna port is determined by the configuration of the adopted base station antenna array, and finally the system is maximized. The purpose of throughput.
The method for designing a three-dimensional beamforming in a multi-user 3D-MIMO system according to claim 1, wherein in step 1), the application is proposed to distribute the user in a new 3D scene in the 3GPP standard, and calculate the user straight. The probability density function of the apex angle of the path of view.
The method for designing a three-dimensional beamforming in a multi-user 3D-MIMO system according to claim 1, wherein in the step 1), the multi-user scheduling algorithm in the system should satisfy the following conditions:

The difference between the apex angles of the users who are simultaneously scheduled at each time is θ d , and the expectation that θ d should be satisfied is as large as possible, that is,

MaxΕθ d (1)

Where E() indicates that the mathematical expectation is taken, and at the same time, all the θ d should be made as close as possible, that is,

among them,
Represents the variance of all θ d .
The method for designing a three-dimensional beamforming in a multi-user 3D-MIMO system according to claim 1, wherein the step 2) specifically includes the following steps:

2-1) The base station selects the simultaneously scheduled users at each time interval using the designed multi-user scheduling algorithm;

2-2) The base station uses a rectangular antenna array consisting of M×N antenna elements, one column of each antenna array as one A separate antenna port that serves one of the users simultaneously scheduled;

2-3) The precoding ω m used by the mth antenna element in each antenna port is

Where M is the number of rows of the antenna array, λ is the carrier wavelength, m is the serial number of the antenna element, d v is the line spacing of the array, and θ etilt is the electronic downtilt angle used.
π is the pi, the electronic downtilt angles used in the precoding used by all antenna elements in the same antenna port are the same, and the electronic downtilt angles in the precoding used by different ports are different.
The method for designing a three-dimensional beamforming in a multi-user 3D-MIMO system according to claim 4, wherein the step 3) specifically includes the following steps:

3-1) The antenna array used by the base station is a rectangular antenna array composed of M×N antenna elements, and the M in the antenna array configuration should satisfy:

among them,
Indicates rounding down, θ M is the only solution that satisfies the following transcendental equation:

among them,

Ln() means taking the natural logarithm, E() means the mathematical expectation, and θ d is the difference of the apex angle of the user who is scheduled at the same time;

3-2) The electronic downtilt angle θ etilt in the precoding used by all antenna elements in the same antenna port is equal and should satisfy

θ etilt =θ z +Δθ (5)

Where θ z is the apex angle of the antenna port serving user, and Δθ is the adjustment angle, which should be satisfied

Δθ=θ M -θ d (6)

Where θ d is the difference between the apex angles of the users scheduled at the same time, and θ M is calculated by the following formula:

Where λ is the carrier wavelength, d v is the row spacing of the array, and M is the number of rows of the antenna array.