CN112543043A - Beam space distributed power distribution method based on non-orthogonal multiple access technology - Google Patents

Abstract

A distributed power allocation method under beam space based on non-orthogonal multiple access technology belongs to the field of wireless communication. In order to solve the problem that the number of user terminals is often larger than the number of radio frequency links, the interference between users is difficult to suppress and the frequency spectrum is caused. Efficiency is significantly reduced. The invention randomly generates the original channel matrix; according to the characteristics of the discrete lens array, the original channel matrix is converted into the beam space; the maximum amplitude beam selection algorithm is used for beam selection, and the beam set to be used by the base station transmitting end is determined, and obtained from the beam set The actual channel matrix between the transmitter and the receiver; the zero-breaking precoding algorithm is used to suppress the interference between different beams; the user cluster is formed, and the non-orthogonal multiple access technology is introduced among the users of the same user cluster; Multiple users of the link perform distributed power allocation between clusters and within clusters to obtain power allocation results. The beneficial effect is that the spectral efficiency of the system is significantly improved.

Description

A distributed power allocation method in beam space based on non-orthogonal multiple access technology

technical field

The invention belongs to the field of wireless communication, and in particular relates to a user power allocation technology in beam space.

Background technique

With the widespread use of mobile devices and the rapid growth of wireless data services, people's demand for wireless network terminal access and communication rates is also increasing; in 5G communications, millimeter-wave massive MIMO beamforming technology is improving It shows great advantages in spectrum utilization, control of interference between users, etc.

After combining massive MIMO beamforming technology with the concept of beam space, the analog-digital hybrid structure combining digital precoding and beam selection algorithm is adopted at the transmitting end, which can significantly reduce the number of radio frequency links and reduce hardware complexity and power consumption. The purpose is to use the discrete lens array (Discrete Lens Array, DLA) to convert the physical space MIMO channel to the beam space, and replace the physical space MIMO with beam space MIMO (Beamspace MIMO, B-MIMO). At this time, each radio frequency link No longer corresponds to a transmitter antenna, but corresponds to a beam in a certain direction; taking the downlink as an example, different single-antenna mobile users meet their own communication needs by selecting a beam in a specific direction; therefore , using the sparse characteristics of the millimeter wave channel in the beam space, the transmitter can select the beam that has a greater effect on the user through the beam selection algorithm to communicate, so as to reduce the number of radio frequency links.

There is also a key problem in the practical process of B-MIMO: the number of users in the system cannot exceed the number of radio frequency links; this is because the degree of freedom provided by the radio frequency link at the transmitter must be greater than or equal to the degree of freedom required by users, otherwise Inter-user interference will be difficult to suppress, resulting in a significant reduction in spectral efficiency; however, in practice, the number of user terminals is often random, and often larger than the number of radio links, especially in 5G mMTC service scenarios.

SUMMARY OF THE INVENTION

The purpose of the present invention is to solve the problem that the number of user terminals is often larger than the number of radio frequency links, which makes it difficult to suppress the interference between users and causes the spectral efficiency to be significantly reduced, and proposes a beam based on non-orthogonal multiple access technology. Distributed power allocation method in space.

A method for distributed power allocation in beam space based on non-orthogonal multiple access technology according to the present invention, the allocation method includes the following steps:

Step 1. According to the millimeter wave channel model, randomly generate the physical space downlink original channel matrix between the base station and multiple single-antenna mobile users;

Step 2: Convert the original channel matrix randomly generated in step 1 to the beam space according to the characteristics of the discrete lens array to obtain the beam space channel matrix;

Step 3: According to the beam space channel matrix obtained in Step 2, use the maximum amplitude beam selection algorithm to select the beam, determine the beam set to be used by the base station transmitter, and obtain the actual channel between the transmitter and the receiver from the beam set. matrix;

Step 4: According to the result of beam selection in Step 3, use a zero-breaking precoding algorithm to suppress interference between different beams; pair two users who select the same beam to form a user cluster, and among users in the same user cluster Introduce non-orthogonal multiple access technology to allow multiple users in the same beam to share time-frequency resources;

Step 5: According to the channel information in the beam space, perform distributed power allocation between clusters and within clusters for multiple users of the downlink, and obtain a power allocation result.

The beneficial effect of the present invention is that based on the advantages of beam selection algorithm and non-orthogonal multiple access (NOMA) technology in the beam space model, the number of radio frequency links is reduced and the number of user access is increased. The distributed power allocation scheme under the beam space of the NOMA technology further improves the system spectrum efficiency; the distributed power allocation scheme under the beam space based on the non-orthogonal multiple access (NOMA) technology proposed by the present invention is On the basis of the beam selection algorithm, the problem that the number of access users in the existing model cannot be greater than the number of radio frequency links is solved by introducing the non-orthogonal multiple access (NOMA) technology; After the (NOMA) technology, a distributed power allocation scheme is adopted, which can ensure the fairness between users to a certain extent and significantly improve the system spectrum efficiency.

Description of drawings

1 is a flowchart of a method for distributed power allocation in beam space based on non-orthogonal multiple access technology according to Embodiment 1;

Fig. 2 shows the average power allocation scheme without introducing the non-orthogonal multiple access technology and the non-orthogonal multiple access technology-based average power allocation scheme when the signal-to-noise ratio is 10dB and each user selects one beam in the first embodiment. Schematic diagram of the spectral efficiency of the distributed power allocation scheme in the beam space, and the curve comparison with the number of users;

Figure 3 shows the average power allocation scheme without introducing the non-orthogonal multiple access technology and the non-orthogonal multiple access-based average power allocation scheme in the first embodiment when the system user access quantity is 30 and each user selects one beam The spectral efficiency of the distributed power allocation scheme in the beam space of the technology, and the curve comparison diagram of the transmission signal-to-noise ratio.

Detailed ways

Embodiment 1: This embodiment is described with reference to FIG. 1 to FIG. 3 . A method for distributed power allocation under beam space based on non-orthogonal multiple access technology described in this embodiment includes the following steps:

Step 1. According to the millimeter wave channel model, randomly generate the physical space downlink original channel matrix between the base station and multiple single-antenna mobile users;

Step 2: Convert the original channel matrix randomly generated in step 1 to the beam space according to the characteristics of the discrete lens array to obtain the beam space channel matrix;

Step 3: According to the beam space channel matrix obtained in Step 2, use the maximum amplitude beam selection algorithm to select the beam, determine the beam set to be used by the base station transmitter, and obtain the actual channel between the transmitter and the receiver from the beam set. matrix;

Step 4: According to the result of beam selection in Step 3, use a zero-breaking precoding algorithm to suppress interference between different beams; pair two users who select the same beam to form a user cluster, and among users in the same user cluster Introduce non-orthogonal multiple access technology to allow multiple users in the same beam to share time-frequency resources;

Step 5: According to the channel information in the beam space, perform distributed power allocation between clusters and within clusters for multiple users of the downlink, and obtain a power allocation result.

In this embodiment, according to the millimeter wave channel model in step 1, the specific method for randomly generating the physical space downlink original channel matrix between the base station and multiple single-antenna mobile users is as follows:

Step 11: Determine the received signal vector; specifically: the number of antennas at the transmitting end of the MIMO system model is N, and N is a positive integer; the receiving end has K single-antenna users, and K is a positive integer; the received signal vector y=[y ₁ ,y ₂ ,…,y _K ] ^T is expressed as

y=H ^H Gs+n (1)

Among them, s=[s ₁ , s ₂ ,...,s _K ] ^T is a K×1-dimensional transmit signal vector, and satisfies E(ss ^H )=I _K , s ^H represents the transmit signal vector s=[s ₁ , s ₂ ,...,s _K ] The conjugate transpose of ^T , E() is mean coincidence, I _K represents a K-dimensional identity matrix; the dimension of the precoding matrix G=[g ₁ ,g ₂ ,...,g _K ] is N×K; n represents noise, that is, cyclic symmetric Gaussian noise with zero mean and variance 1; H=[h ₁ , h ₂ ,...,h _K ] is a physical channel matrix of dimension N×K, where each column h _k represents the channel vector between the base station and mobile user k, and the dimension is N×1;

Steps 1 and 2: Determine the original channel matrix according to the millimeter wave channel model; specifically: according to the millimeter wave channel model, it can be known that the original channel matrix includes the base station and the user k;

The channel vector between the base station and user k is expressed as:

Among them, β is the path loss, θ is the path angle, θ _k,i represents the complex path angle corresponding to the different paths of the k-th user, θ _k,0 represents the line-of-sight path angle corresponding to the different paths of the k-th user, β _{k, i} represents the complex path loss corresponding to the different paths of the k-th user, β _k,0 represents the line-of-sight path loss corresponding to the different paths of the k-th user, and the magnitude of the multipath component |β _k,i | ) component |β _k,0 | is 5 to 10 dB smaller; N _p represents the number of multipaths, and a _N is an N×1-dimensional control vector, expressed as:

a _N =[e ^-j2πθi ] i∈Γ _(N)

Among them, Γ(N)={l-(N-1)/2:l=0,1,...,N-1} is a symmetric set centered at 0, and l represents each element in the symmetric set And set a variable;

The path angle θ is expressed as:

Among them, λ represents the signal wavelength, d represents the antenna spacing; d=λ/2 is the sample spacing of the antenna aperture domain, and the direction angle φ∈[-π/2, π/2] in the physical space, then θ∈[-1 /2,1/2].

因此矩阵U可表示为In this embodiment, the specific method for obtaining the beam space channel matrix in step 2 is: transform the physical space channel matrix into the beam space to obtain the specific expression of the matrix U; according to the characteristics of the discrete lens array, the columns of the matrix U The vector corresponds to n fixed spatial frequency/angle control vectors, and each vector has a fixed spacing

So the matrix U can be expressed as

Transforming the physical channel into the beam space, the MIMO system in the beam space can be expressed as

表示的是K个用户的发射功率，满足

即发射功率总和不能超过基站端最大的发射功率。where H _b is the channel matrix in the beam space, and satisfies H _b =U ^H H; G _b is the precoding matrix in the beam space and satisfies G _b =UG; P=diag{p} is a diagonal matrix, which diagonal element

Represents the transmit power of K users, satisfying

That is, the total transmit power cannot exceed the maximum transmit power of the base station.

In this embodiment, the specific method of using the maximum amplitude beam selection algorithm for beam selection in step 3 is as follows:

Select several beams with larger amplitudes. In order to describe the algorithm mathematically, define a set

为H_b的第i行k列元素；M^(k)是针对于第k个用户的选择集合；ξ^(k)∈[0,1]为选择门限，通过调整该值可以选择不同数量的主要波束，为了使得每个用户都选择到最少的波束，ξ^(k)对于每个用户必须是独立取值的。in,

is the i-th row and k-column elements of H _b ; M ^(k) is the selection set for the k-th user; ξ ^(k) ∈ [0,1] is the selection threshold, and different numbers of main Beam, in order to make each user select the least beam, ξ ^(k) must be independently valued for each user.

In this embodiment, the actual beam space channel matrix is obtained according to the beam selection result, and it can be known from the MMS beam selection algorithm that there may be a single user or multiple users in the same beam. If there is a single user in the beam, then its vector in the beam space represents a vector of the actual beam space channel matrix; if there are several users in the beam, then we need to use other methods to determine the beam in the actual beam space channel form in the matrix.

The specific method for obtaining the actual channel matrix between the transmitter and the receiver from the beam set in step 3 is as follows:

If there are multiple users in the same beam, and NOMA technology is used among multiple users, the corresponding vector of the beam in the actual beam channel matrix is obtained according to the singular value decomposition method. The specific process is as follows:

进行奇异值分解，可得：Among them, H _m is the channel matrix composed of all user channel vectors in the mth beam, the dimension is N _RF ×|S _m |, and |S _m | is the number of users in the beam; H _m consists of |S _m | composed of column vectors, h _i,m is the channel vector of the i-th user in the m-th beam;

Perform singular value decomposition to get:

Among them, U _m is the left singular value decomposition matrix of |S _m |×|S _m | dimension, ∑ _m is the diagonal matrix, and V _m is the right singular value decomposition matrix of N _RF ×N _RF dimension;

The equivalent channel vector of the mth beam can be obtained by:

为

最大奇异值所对应的左奇异分解矩阵的列向量，

则表示信道矩阵

中的一个波束向量；in,

for

the column vector of the left singular decomposition matrix corresponding to the largest singular value,

then represents the channel matrix

a beam vector in ;

可表示为：According to the principle of zero-breaking precoding, the precoding matrix

can be expressed as:

表示预编码矩阵

的第n个列向量，

表示经过波束选择后的信道矩阵，

则表示

的共轭转置。in,

represents the precoding matrix

the nth column vector of ,

represents the channel matrix after beam selection,

means

The conjugate transpose of .

In this embodiment, the specific process of performing distributed power allocation between clusters and within clusters to multiple downlink users in step 5 is as follows:

Step 51: Perform power allocation between user clusters in units of user clusters;

Step 52: Perform power allocation within user clusters in units of users.

In this embodiment, the specific criteria for power allocation between user clusters in the unit of user clusters in step 51 are:

The total power of a single-user cluster is P/K, and the total power of a dual-user cluster is (2P)/K; where K is the total number of downlink single-antenna users, and P is the total transmit power of the base station.

In this embodiment, the specific criteria for power allocation within user clusters in units of users in step 52 are: the total power of the user clusters is P/K; the dual-user clusters use a fractional-order power allocation scheme, and at this time the first The transmit power of i dual-user clusters can be expressed as:

Among them, α ^ftpa ∈ [0,1] is the power allocation factor, _gi (j) is the equivalent channel gain, and _ni (j) is the noise and interference.

In this embodiment, according to the power allocation result in step 5, the method of spectral efficiency performance comparison is used for verification, and the system performance using the average power allocation scheme is compared to verify the beam space down distribution based on the non-orthogonal multiple access technology. The beneficial effect of the power distribution scheme of the present embodiment; the specific process of preparation in this embodiment is:

The simulation conditions are: millimeter wave channel model (S-V channel model), cyclic symmetric Gaussian noise with zero mean and variance 1, the number of antennas at the base station transmitter is 81, the user receiver is equipped with a single receiver antenna, and users in the same beam The maximum number is two; the millimeter-wave channel model only considers time-invariant channels, the number of multipath bars is 2, and the Rice factor is set to 5, these parameters are in good agreement with the real channel conditions;

On the basis of the above conditions, the system performance of the distributed power allocation scheme in the beam space based on the non-orthogonal multiple access technology under different conditions is verified by simulation.

It can be seen from Figure 2 that when the transmit signal-to-noise ratio is 10dB and each user selects one beam, when the number of users increases to 30, the average power allocation scheme without introducing the non-orthogonal multiple access technology is adopted When the number of users increases, the spectral efficiency of the system decreases significantly; and when the distributed power allocation scheme in the beam space based on the non-orthogonal multiple access technology is adopted, the spectral efficiency increases with the increase of the number of users, and always remains constant. former.

It can be seen from Figure 2 that under the same signal-to-noise ratio, the distributed power allocation scheme in the beam space based on the non-orthogonal multiple access technology can increase the user access of the massive MIMO system and improve the spectral efficiency of the system.

As can be seen from Figure 3, in the case where the number of users in the system is 30 and each user selects one beam, the average power allocation scheme that does not introduce the non-orthogonal multiple access technology is different from that based on the non-orthogonal multiple access technology. The spectral efficiency of the distributed power allocation scheme in the beam space of the technology will increase with the increase of the transmit signal-to-noise ratio, and the performance of the latter is always better than that of the former.

It can be further verified from FIG. 3 that in the case of the same system user access volume, the distributed power allocation scheme in the beam space based on the non-orthogonal multiple access technology can improve the spectral efficiency of the system.

The present invention can also have other various embodiments. Without departing from the spirit and essence of the present invention, those skilled in the art can make various corresponding changes and deformations according to the present invention, but these corresponding changes and deformations are all It should belong to the protection scope of the appended claims of the present invention.

Claims (8)

1. A distributed power allocation method in beam space based on non-orthogonal multiple access technology, characterized in that the allocation method comprises the following steps: Step 1. According to the millimeter wave channel model, randomly generate the physical space downlink original channel matrix between the base station and multiple single-antenna mobile users; Step 2: Convert the original channel matrix randomly generated in step 1 to the beam space according to the characteristics of the discrete lens array to obtain the beam space channel matrix; Step 3: According to the beam space channel matrix obtained in Step 2, use the maximum amplitude beam selection algorithm to select the beam, determine the beam set to be used by the base station transmitter, and obtain the actual channel between the transmitter and the receiver from the beam set. matrix; Step 4: According to the result of beam selection in Step 3, use a zero-breaking precoding algorithm to suppress interference between different beams; pair two users who select the same beam to form a user cluster, and among users in the same user cluster Introduce non-orthogonal multiple access technology to allow multiple users in the same beam to share time-frequency resources; Step 5: According to the channel information in the beam space, perform distributed power allocation between clusters and within clusters for multiple users of the downlink, and obtain a power allocation result. 2. a kind of distributed power allocation method under beam space based on non-orthogonal multiple access technology according to claim 1, it is characterized in that, in step 1, according to millimeter wave channel model, randomly generate base station and a plurality of single The specific method of the physical space downlink original channel matrix between the antenna mobile users is: Step 11: Determine the received signal vector; specifically: the number of antennas at the transmitting end of the MIMO system model is N, and N is a positive integer; the receiving end has K single-antenna users, and K is a positive integer; the received signal vector y=[y ₁ ,y ₂ ,…,y _K ] ^T is expressed as y=H ^H Gs+n (1) Among them, s=[s ₁ , s ₂ ,...,s _K ] ^T is a K×1-dimensional transmit signal vector, and satisfies E(ss ^H )=I _K , s ^H represents the transmit signal vector s=[s ₁ , s ₂ ,...,s _K ] The conjugate transpose of ^T , E() is mean coincidence, I _K represents a K-dimensional identity matrix; the dimension of the precoding matrix G=[g ₁ ,g ₂ ,...,g _K ] is N×K; n represents noise, that is, cyclic symmetric Gaussian noise with zero mean and variance 1; H=[h ₁ , h ₂ ,...,h _K ] is a physical channel matrix of dimension N×K, where each column h _k represents the channel vector between the base station and mobile user k, and the dimension is N×1; Steps 1 and 2: Determine the original channel matrix according to the millimeter wave channel model; specifically: according to the millimeter wave channel model, it can be known that the original channel matrix includes the base station and the user k; The channel vector between the base station and user k is expressed as:

Among them, β is the path loss, θ is the path angle, θ _k,i represents the complex path angle corresponding to the different paths of the kth user, θ _k,0 represents the line-of-sight path angle β _k,i corresponding to the different paths of the kth user represents the complex path loss corresponding to the different paths of the kth user, β _k,0 represents the line-of-sight path loss corresponding to the different paths of the kth user, and the magnitude of the multipath component |β _k,i | is greater than the line-of-sight transmission path (LoS) The component |β _k,0 | is 5 to 10 dB smaller; N _p represents the number of multipaths, and a _N is an N × 1-dimensional control vector expressed as: a _N =[e ^-j2πθi ] i∈Γ _(N) Among them, Γ(N)={l-(N-1)/2:l=0,1,...,N-1} is a symmetric set centered at 0, and l represents each element in the symmetric set And set a variable; The path angle θ is expressed as:

Among them, λ represents the signal wavelength, d represents the antenna spacing; d=λ/2 is the sample spacing of the antenna aperture domain, and the direction angle φ∈[-π/2, π/2] in the physical space, then θ∈[-1 /2,1/2]. 3. a kind of distributed power allocation method under beam space based on non-orthogonal multiple access technology according to claim 2, is characterized in that, the concrete method that obtains beam space channel matrix in step 2 is: The channel matrix is transformed into the beam space, and the specific expression of the matrix U is obtained; according to the characteristics of the discrete lens array, the column vectors of the matrix U correspond to n fixed spatial frequency/angle control vectors, and each vector has a fixed spacing

So the matrix U can be expressed as

Transforming the physical channel into the beam space, the MIMO system in the beam space can be expressed as

where H _b is the channel matrix in the beam space, and satisfies H _b =U ^H H; G _b is the precoding matrix in the beam space and satisfies G _b =UG; P=diag{p} is a diagonal matrix, which diagonal element

Represents the transmit power of K users, satisfying

That is, the total transmit power cannot exceed the maximum transmit power of the base station. 4. a kind of distributed power allocation method in beam space based on non-orthogonal multiple access technology according to claim 3, is characterized in that, adopts the concrete method of beam selection of maximum amplitude beam selection algorithm in step 3 for: Select several beams with larger amplitudes. In order to describe the algorithm mathematically, define a set

in,

is the i-th row and k-column elements of H _b ; M ^(k) is the selection set for the k-th user; ξ ^(k) ∈ [0,1] is the selection threshold, and different numbers of main Beam, in order to make each user select the least beam, ξ ^(k) must be independently valued for each user. 5. a kind of distributed power allocation method under the beam space based on non-orthogonal multiple access technology according to claim 4, is characterized in that, in step 3, obtains the actual between the transmitter and the receiver by the beam set The specific method of the channel matrix is: If there are multiple users in the same beam, and NOMA technology is used among multiple users, the corresponding vector of the beam in the actual beam channel matrix is obtained according to the singular value decomposition method. The specific process is as follows:

Among them, H _m is the channel matrix composed of all user channel vectors in the mth beam, the dimension is N _RF ×|S _m |, and |S _m | is the number of users in the beam; H _m consists of |S _m | composed of column vectors, h _i,m is the channel vector of the i-th user in the m-th beam;

Perform singular value decomposition to get:

Among them, U _m is the left singular value decomposition matrix of |S _m |×|S _m | dimension, ∑ _m is the diagonal matrix, and V _m is the right singular value decomposition matrix of N _RF ×N _RF dimension; The equivalent channel vector of the mth beam can be obtained by:

in,

for

the column vector of the left singular decomposition matrix corresponding to the largest singular value,

then represents the channel matrix

a beam vector in ; According to the principle of zero-breaking precoding, the precoding matrix

can be expressed as:

in,

represents the precoding matrix

the nth column vector of ,

represents the channel matrix after beam selection,

express

The conjugate transpose of . 6. a kind of distributed power allocation method under beam space based on non-orthogonal multiple access technology according to claim 5, it is characterized in that, in step 5, perform inter-cluster and cluster to a plurality of users of downlink The specific process of internal distributed power distribution is as follows: Step 51: Perform power allocation between user clusters in units of user clusters; Step 52: Perform power allocation within user clusters in units of users. 7. A kind of distributed power allocation method under beam space based on non-orthogonal multiple access technology according to claim 6, it is characterized in that, in described step 51, use user cluster as unit to carry out power between user clusters The specific criteria for allocation are: The total power of a single-user cluster is P/K, and the total power of a dual-user cluster is (2P)/K; where K is the total number of downlink single-antenna users, and P is the total transmit power of the base station. 8. The method for distributed power allocation under a beam space based on a non-orthogonal multiple access technology according to claim 6, wherein in the step 52, power allocation within user clusters is performed in units of users The specific criteria are: the total power of the user cluster is P/K; the dual-user cluster uses a fractional-order power allocation scheme, and the transmit power of the i-th dual-user cluster can be expressed as:

Among them, α ^ftpa ∈ [0,1] is the power allocation factor, _gi (j) is the equivalent channel gain, and _ni (j) is the noise and interference.

Images