CN113747558B

CN113747558B - Power control method of MISO-NOMA uplink channel

Info

Publication number: CN113747558B
Application number: CN202110973495.6A
Authority: CN
Inventors: 刘铭; 张军霞; 熊轲; 程慈航; 张天壮; 王鑫; 徐宇轩
Original assignee: Beijing Jiaotong University
Current assignee: Beijing Jiaotong University
Priority date: 2021-08-24
Filing date: 2021-08-24
Publication date: 2022-12-06
Anticipated expiration: 2041-08-24
Also published as: CN113747558A

Abstract

The invention relates to a power control method of a MISO-NOMA uplink channel, which comprises the following steps: step 1: establishing an uplink transmission scene containing intra-group and inter-group interference in the MISO-NOMA system; step 2: the base station performs equalization processing on the received signals; and step 3: an uplink NOMA user grouping scheme based on K-means is formulated; and 4, step 4: after NOMA user grouping is completed, a user demodulation sequence capable of minimizing the transmission power and the minimum uplink transmission power is given by considering the minimum uplink transmission power closed-form solution of users in a group constrained by the minimum data rate; and 5: carrying out user demodulation in the same NOMA transmission group by using SIC; step 6: after the minimum transmitting power of the users in the group and the user demodulation sequence are obtained, the iterative power control method for calculating the inter-group power based on the strength of inter-group interference realizes the minimization of the total transmitting power of the system under the condition of determining the NOMA transmission group.

Description

Power control method of MISO-NOMA uplink channel

Technical Field

The invention belongs to the technical field of wireless communication, and particularly relates to a power control method of a MISO-NOMA uplink channel.

Background

In the current wireless network development, the contradiction between the massive access requirements and the limited spectrum resources is more and more prominent. How to support large-scale access requirements with limited resources has become an urgent problem to be solved in mobile communication networks. On the other hand, a more reasonable resource control method needs to be designed to reduce the energy consumption of the wireless network. As shown in fig. 1 and fig. 2, a Non-Orthogonal Multiple Access (NOMA) technique of power multiplexing is adopted, and by transmitting end multi-user superposition transmission and receiving end Serial Interference Cancellation (SIC), multiplexing of time-frequency resources can be realized to meet the requirement of large-scale wireless Access. In addition, a large-scale Multiple-Input Multiple-Output (MIMO) technology introduces wider spatial freedom through the deployment of a large number of antennas, reduces inter-user interference, and further improves the utilization efficiency of frequency spectrum resources, the system capacity and the signal transmission reliability. Therefore, it is an effective means to support access to a large number of users in the uplink communication link by using an access technology combining massive MIMO with NOMA and realizing efficient utilization of time-frequency resources.

The NOMA technology realizes multi-user superposition transmission, which effectively improves the utilization rate of wireless resources, but simultaneously causes the increase of interference in the NOMA system. The level of interference limits the performance of the overall system. Power control is an effective way to solve the problem of interference in NOMA systems. Meanwhile, through reasonable adjustment of the transmission power of the user equipment, when the user signals using the wireless resources on the same time-frequency reach the receiver, a remarkable power difference can be formed. Furthermore, the receiver can demodulate different user signals one by a Successive Interference Cancellation (SIC) method, thereby implementing multi-user detection. Therefore, the reasonable power control in the NOMA system can realize the effective multiplexing of radio resources and the reliable demodulation of a plurality of user signals.

The invention provides an uplink transmission power control technology of a MISO-NOMA system aiming at the application requirements of a future wireless network and aiming at improving the user access quantity and reducing the energy consumption. Unlike the prior art, the present invention considers both intra-group interference and inter-group interference. First, a user grouping scheme capable of reducing inter-group interference is proposed using antenna correlation characteristics of the MIMO system. Secondly, an iterative NOMA transmission group power control algorithm is provided, so that the minimization of the total transmission power of the system based on the user grouping result is realized, and the total power consumption of the system is reduced.

Disclosure of Invention

In most existing researches related to uplink resource optimization of the MISO-NOMA system, the transmission rate under the condition of maximizing unit energy consumption is generally considered, and the minimum transmission rate requirement of a user is ignored. In practice, especially under a massive uplink access scenario oriented to future mobile communication, how to reduce energy consumption as much as possible while meeting a user rate requirement is still a problem to be solved urgently. Secondly, in a multi-user NOMA system, a user is interfered by not only other users in the same NOMA transmission group (referred to as intra-group interference for short), but also users in other NOMA transmission groups (referred to as inter-group interference for short). Most of the current studies only consider reducing intra-group interference, and neglect the study of inter-group interference suppression methods. Research shows that inter-group interference has certain influence on user power control, which results in limited performance improvement of uplink transmission.

Aiming at the defects in the prior art, the invention aims to provide a power control method of an MISO-NOMA uplink channel.

In order to achieve the purpose, the invention adopts the technical scheme that:

a power control method of MISO-NOMA uplink channel includes the following steps:

step 1: establishing an uplink transmission scene of the MISO-NOMA system to obtain a signal y received by the base station;

and 2, step: the base station performs equalization processing on the received signal y through maximum ratio combination, optimizes a user grouping mode in order to reduce the interference among groups, reduces the correlation degree among channel vectors of different groups of users, and obtains

Indicating the ith group to be demodulatedThe sum of the inter-group interference plus noise experienced by the user;

and step 3: an uplink NOMA user grouping scheme based on K-means is formulated, and NOMA user grouping is completed;

and 4, step 4: obtaining a minimum uplink transmitting power closed-form solution required by any user i in the group by considering the minimum data rate constraint of the users in the group;

and 5: obtaining an optimal demodulation sequence, and demodulating users in the same NOMA transmission group by adopting serial interference elimination;

step 6: after the minimum uplink transmission power and the optimal demodulation sequence required by any user i are obtained, the iterative power control method for calculating the inter-group power based on the strength of inter-group interference realizes the minimization of the total transmission power of the system under the condition of determining the NOMA transmission group.

On the basis of the scheme, the step 1 specifically comprises the following steps:

step 1.1: establishing an uplink transmission scene of a single-cell MISO-NOMA system, wherein a base station simultaneously provides service for K users, K represents user serial number, K =1,2, \8230, K is provided with N antennas, N represents antenna serial number of the base station, N =1,2, \8230, N is provided with one antenna for each user, and assuming that all users in a cell are divided into L NOMA transmission groups, L represents serial number of the NOMA transmission group, L =1,2, \8230, L is less than N; defining a correlation matrix G with the size of K multiplied by K, which is used for expressing the correlation of channel vectors between a base station and K users;

step 1.2: performing Cholesky decomposition on the correlation matrix G to obtain:

wherein,

representing a conjugate transpose for a triangular matrix obtained after Cholesky decomposition;

establishing a channel matrix H, H = [ H ] between a base station and a user ₁ ,h ₂ ,…,h _K ] ^T The size is K N, H is expressed as:

wherein, V is a matrix with the size of K multiplied by N, and comprises wireless channel vectors between K users and N base station antennas;

the channel vectors corresponding to different users are independent of each other, and each element V in the matrix V _k,n Can be decomposed into path loss beta _k,n And small scale fading ξ _k,n Where k and n are subscripts of the element in the matrix V, in particular V _k,n Expressed as:

due to the close distance between different base station antennas, the path loss beta experienced by the signal of the same user reaching different base station antennas can be considered _k,n Equal, small scale fading ξ _k,n Are independently and identically distributed;

the correlation matrix G is a symmetric matrix satisfying:

wherein for the element g of the ith row and the jth column in the matrix _i,j And element g of jth row and ith column _j,i Has g _i,j ＝g _j,i And when i = j, g _i,j ＝1。

Each element G of the correlation matrix G _i,j The correlation between the channels of the user i and the user j is represented, and is calculated by an equation (5):

wherein (x) _i ,y _i ) And (x) _j ,y _j ) The representation being the position coordinates of two users, λ _c Denotes the correlation distance, λ _c The smaller the value of (c), the stronger the correlation between users in the cell.

The radio channel vector between the base station and the kth user is represented as:

h _k ＝[h _k,1 ,h _k,2 ,...,h _k,n ,...h _k,N ] ^T (6)

wherein the base station is provided with N antennas, so that h is used _k,n Corresponding to a channel vector between the kth user and the nth antenna of the base station;

let p be _k Representing the uplink transmit power, s, of the k-th user _k Is the complex value data symbol sent by the kth user, z is the complex value vector of nx 1, representing additive white gaussian noise, the signal received by the base station is:

on the basis of the scheme, the step 2 specifically comprises the following steps:

step 2.1: the base station performs equalization of signals through maximum ratio combining, and an equalization matrix is expressed as:

W＝H ^* (8)

wherein, H is a channel matrix between the base station and the user;

the signal after maximum ratio combining equalization is expressed as:

wherein, the integral and the | | respectively represent the conjugate transpose and the vector norm; n is the number of antennas of the base station,

is a channel vector h _k The conjugate transpose of (1); y isA signal received by a base station;

is the k-th user signal weighted by a real-valued factor, where p _k Is the uplink transmit power, h, of the kth user _k Is a wireless channel vector between the base station and the kth user; s is _k A complex-valued data symbol transmitted for the kth user;

including intra-group interference and inter-group interference, k' is a subscript index indicating users other than the k-th user;

is the noise component after equalization, z is additive white gaussian noise;

step 2.2: in order to reduce the inter-group interference, it is necessary to optimize the user grouping method and reduce the correlation between the channel vectors of different groups of users, assuming that there is M in the ith group _l Individual users, defining a vector pi _l ＝[π _l (1),π _l (2),…,π _l (M _l )]For the base station demodulation order of the user signals of the l-th group, where _l (i) Denotes an index of the i-th demodulated user when demodulating the l-th group of user signals of the signal by using the SIC.

Step 2.3: the base station firstly demodulates the subscript pi in the first group _l (1) To pi _l (i-1), then removing these signal components from the received signal, and then demodulating the signal indexed by π _l (i) When the index in the l-th group is pi _l (i + 1) to π _l (M _l ) Is an interference component.

Use of

Represents the sum of the intergroup interference plus noise experienced by the ith demodulated user in the ith group:

wherein M is _l' The number of users in the l' th group;

the uplink transmission power of the kth user in the ith user group is obtained;

the conjugate transpose of the channel vector for the ith demodulation user in the ith group;

channel vectors for the kth 'demodulated user in the l' user group;

for the index in the ith user group of pi _l (i) The noise received by the user; and N is the number of antennas of the base station.

On the basis of the scheme, the step 3 specifically comprises the following steps:

step 3.1: based on the optimal grouping number of SUS and the selection algorithm of the clustering center, the optimal grouping number L and the initial clustering center C corresponding to each group are calculated ⁰ 。

Step 3.2: and sequentially calculating the distance between each user to be grouped and each group of initial clustering centers, dividing the users into L groups, updating the clustering centers, and regrouping until the clustering centers are not changed any more.

On the basis of the above scheme, step 3.1 specifically includes:

the SUS-based optimal grouping number and clustering center selection algorithm specifically comprises the following steps:

step 3.1.1: according to the user channel matrix H = [ H = ₁ ,h ₂ ,…,h _K ] ^T Setting a semi-orthogonality factor alpha, wherein alpha is a constant between 0 and 1 and represents orthogonality among users;

step 3.1.2: initializing a candidate user set gamma, gamma = {1, \8230;, K }, wherein the candidate user set gamma represents all users which are not selected into the cluster center set, and making the cluster center set C as an empty set, the optimal grouping number L equal to 1 and the iteration index i equal to 1.

Step 3.1.3: calculating a projection vector of each user k in the candidate user set gamma to a null space of a selected clustering center, wherein the projection vector of the user k to the null space of the selected clustering center is expressed as:

wherein h is _k Representing a radio channel vector between the base station and the kth user;

step 3.1.4: performing N iterations, traversing the candidate user set in each iteration, and calculating g of each user _k And according to g _k The maximum principle identifies the index of the cluster center:

wherein,

selecting subscript index corresponding to the clustering center;

step 3.1.5: updating a clustering center set S:

step 3.1.6: updating the candidate user set according to the semi-orthogonality of the users in the candidate user set gamma and the clustering center set S so as to ensure that the users in the candidate user set are semi-orthogonal to the clustering center set, wherein the semi-orthogonality is determined by alpha:

users which do not meet the semi-orthogonal requirement in the candidate user set gamma are directly discarded;

step 3.1.7: updating the optimal group number sum and the iteration index obtained by the current iteration round number, L ← L +1, i ← i + 1';

step 3.1.8: repeating the step 3.1.3-3.1.7 until the current iteration index value i is more than or equal to N or the candidate user set is an empty set;

step 3.1.9: finally, the initial clustering center C is obtained ⁰ ，

And an optimal number of packets L.

On the basis of the above scheme, step 3.2 specifically includes:

step 3.2.1: according to the user channel matrix H = [ H = ₁ ,h ₂ ,…,h _K ] ^T Setting a weight factor epsilon and a balance factor lambda; the weight factor epsilon is a weighted value of each distance item in the total distance weight calculated in the grouping process, and the balance factor lambda is a penalty value for balancing the distance between a certain group and a user when the group of users is excessive.

Step 3.2.2: initializing cluster center set C ⁰ Make the cluster center set C ⁰ Is an empty set, and makes a packet matrix mu = mu ₁ ∪μ ₂ …∪μ _L T =0, the current iteration round number t is 0;

step 3.2.3: obtaining the initial clustering center calculation C obtained in the step 3.1.9 ⁰ And the optimal number of packets L;

step 3.2.4: traversing each user in the candidate user set, and calculating the similarity between the user and each cluster center channel vector and the minimum channel gain difference between the user and each group;

using two channel vectors h _i And h _j By the cosine of (c) h _i And h _j The correlation of (a):

where | represents the modulus of the complex number and | represents the norm of the vector.

The squared difference of the normalized channel vectors is used to measure the channel gain difference between users, and the minimum gain difference within a group is used to measure the channel gain difference between user i and the l-th group:

wherein, let | h _max | | is the largest channel gain value among the cell users,

and

respectively normalizing the energy of the channel vectors corresponding to the ith and ith' users; omega _l A subscript set for users included in the l-th group;

step 3.2.5: on the basis of the formula (14) and the formula (15), in consideration of the grouping balance, a distance function is established for calculating the distance Dist from each user k to the l-th group _k,l As shown in equation (16):

wherein ε ∈ (0, 1) represents a weighting factor, when the total number of users in the l-th group M _l When the distance between the subsequent ungrouped user and the cluster center of the group is calculated when a certain threshold value is exceeded, adding a balance factor lambda, wherein lambda > 1; the distance between the subsequent users and the group is increased, and the possibility that the group is divided into more users is reduced, so that the number of users in different groups is balanced;

step 3.2.6: the group to which the user k belongs is selected,

wherein

Represents the index of the grouping closest to user k;

step 3.2.7: after the t-th round of clustering, the cluster center of each group is updated by equation (17):

wherein M is _l The number of users in the ith group is represented, and j represents the index of the subscript of other users except i in the ith group.

Step 3.2.8: update iteration index value, t ← t +1.

Step 3.2.9: repeating the steps 3.2.3-3.2.8, updating the users in the group until new C ^t No change occurs;

step 3.2.10: and outputting the user grouping matrix mu.

On the basis of the above scheme, step 4 specifically includes:

step 4.1: assuming that there are M users in a group, given a demodulation order π _l ＝[π _l (1),π _l (2),…,π _l (M)]Assuming that the data rate constraints of all users in the ith group are equal and satisfy the minimum data rate constraint r:

wherein,

the achievable data rate for the ith demodulated user in the ith group;

defining the target signal to interference plus noise ratio requirement needed to be met by a user to meet the minimum data rate constraint r as gamma:

wherein r is a minimum data rate constraint and W is a system bandwidth;

and calculating the minimum uplink transmission power sum in the group through analysis and derivation based on the minimum data rate constraint r.

Step 4.2: analysing last demodulated user pi _l (M) the minimum data rate constraint to be satisfied is:

wherein,

are respectively the pi-th group _l The sum of the (M) demodulated users' uplink transmit power, channel vector, and inter-group interference plus noise.

Obviously, the user power will be minimized when the constraint in equation (21) reaches the equality condition, so that

Denotes the pi-th group in the l-th group _l The minimum uplink transmitting power of (M) demodulation users is obtained:

wherein, gamma is the target signal-to-interference-and-noise ratio requirement, N is the number of antennas of the base station, l, pi _l (M) is the index of the mth demodulation user of the ith group after grouping,

respectively represent the ith group pi _l (M) channel vectors of demodulated users and inter-group interference plus noise sums,

is a constant related to the target sir requirement, the number of antennas at the base station and the user channels in the group.

Step 4.3: solving in the same way

And will beSubstituting the formula (22) to obtain the minimum uplink transmission power of the M-1 user as:

step 4.4: the process is repeated continuously, and according to the calculation result, the minimum uplink transmission power required by any user i is obtained, as follows:

wherein,

channels of users corresponding to respective demodulation orders

And conjugate transpose thereof

And a constant related to the target SINR signal-to-interference-and-noise ratio gamma, k represents the index of the first j-1 demodulation orders, and k' is the order of the 1 st to the kth demodulation.

On the basis of the above scheme, step 5 specifically includes:

after the minimum uplink transmission power in each group is obtained, the optimal demodulation sequence of the users in the group is given

In comparison to other demodulation sequences,

lower power consumption values can be achieved.

For an arbitrary NOMA transmission group, it is assumed that all users within the group have the same purposeMarking data rate, order

Indicating the user demodulation order. If and only if the user satisfies

The sum of the transmission powers of the groups is minimal, i.e. the demodulation order is then the case, when the users with higher interference demodulate first

For an optimal demodulation order.

On the basis of the above scheme, step 6 specifically includes:

step 6.1: according to the user channel matrix H = [ H = ₁ ,h ₂ ,…,h _K ] ^T And the number of users K, the number of antennas N, and the minimum data rate constraint r for the users.

Step 6.2: initializing power sum vectors q for groups ⁰ Power vector p for each user ⁰ The maximum iteration number maxter of the algorithm, and the current iteration round number t.

Step 6.3: in each iteration, sequentially traversing the users of each group to obtain the number M of users of the first group _l And obtaining the optimal demodulation sequence obtained in step 5

Step 6.4: from the last demodulation, i.e. Mth _l The demodulation users start to calculate in sequence, firstly obtaining

And

wherein

Is in the optimal demodulation order

Next, index of current ith demodulation user in current ith group,

is the index of the j demodulation user that demodulates thereafter.

Step 6.5: calculating the current NOMA transmission group of all users according to formula (27)

And

wherein,

represents the sum of the intergroup interference plus noise of the current ith group of demodulated users,

represents the sum of intergroup interference plus noise of the j-th demodulated user demodulated after i;

wherein,

Channels of users corresponding to respective demodulation orders

And together therewithYoke transpose

And a constant related to the target SINR signal-to-interference-and-noise ratio gamma, wherein k represents the index of the first j-1 demodulation sequences, and k' is the sequence from 1 st to k th demodulation;

and

is the sum of the interference plus noise between the current ith group of demodulated users and the jth user demodulated thereafter, and the value of the sum is equal to the power q of other groups _-l And (6) correlating.

Step 6.6: further, in the current iteration round number, the minimum transmitting power sum q of the l group in the t +1 round iteration is calculated according to the formula (28) _l ^(t+1) ；

Wherein q is _l ^(t+1) Refers to the minimum sum of transmit power, M, of the l-th group in the t +1 th iteration _l Indicating the number of users of the group.

Step 6.7: sequentially calculating the minimum uplink transmitting power of each user i in the ith group in the t +1 th iteration according to a formula (29)

Wherein,

represents the minimum uplink transmission power of the ith demodulated user in the ith group in the t +1 round iteration,

is a constant related to the target sir requirement, the number of antennas in the bs and the user channels in the group.

Step 6.8: update the iteration index value, t ← t +1.

Step 6.9: and 6.3-6.8, and updating the power value of each group and each user in the group until t = MaxIter.

Step 6.10: outputting power sum vector of each group

And minimum uplink transmission power vector of each user

The invention has the beneficial effects that:

1. the invention considers the interference between groups when carrying out user grouping and power control in the groups, and can achieve better power consumption and energy efficiency compared with the traditional scheme;

2. the two-step power control scheme provided by the invention solves the problem of non-convex NOMA power control which is difficult to solve by the traditional mathematical method.

3. The grouping iteration power optimization algorithm provided by the invention realizes the minimization of the system power under the condition of considering the interference.

4. The invention can achieve the effect close to the optimal effect under the lower complexity cost.

Drawings

The invention has the following drawings:

fig. 1 is a schematic diagram of the transmission process of user signals in a transmission group in a MISO-NOMA system.

Fig. 2 is a schematic diagram of a NOMA packet SIC process at a receiving end of a base station.

Fig. 3 is a schematic diagram of a multi-user upstream NOMA system with multiple groups.

Fig. 4 is a schematic diagram of a power consumption variation curve of each method under different user numbers, wherein r =5Mbits/s.

Fig. 5 is a schematic diagram of an energy efficiency change curve of each method under different user numbers, wherein r =5Mbits/s.

Fig. 6 is a schematic diagram of a variation curve of the outage probability under different data rate constraints of the methods, where K =10.

Detailed Description

The present invention will be described in further detail with reference to FIGS. 3 to 6.

step 1: uplink transmission scene for establishing MISO-NOMA system

Step 1.1: consider the uplink transmission scenario of a single cell MISO-NOMA system, as shown in fig. 3. The base station provides service for K users simultaneously, wherein K represents the serial number of the user, K =1,2, \ 8230, K, the base station is provided with N antennas, N represents the serial number of the antenna of the base station, N =1,2, \ 8230, N, each user is provided with one antenna, all users in a cell are assumed to be divided into L NOMA transmission groups, L represents the serial number of the NOMA transmission group, and L =1,2, \ 8230, L is less than N; defining a correlation matrix G with the size of K multiplied by K, which is used for expressing the correlation of channel vectors between a base station and K users;

wherein,

for the triangular matrix obtained after Cholesky decomposition, the conjugate transpose of the matrix is represented.

the channel vectors corresponding to different users are independent of each other, and each element V in the matrix V _k,n Can be decomposed into path loss beta _k,n And small scale fading ξ _k,n Where k and n are subscripts of the element in the matrix V. In particular, v _k,n Expressed as:

the correlation matrix G is a symmetric matrix satisfying:

wherein, for the element g of the ith row and the jth column in the matrix _i,j And element g of jth row and ith column _j,i Has g _i,j ＝g _j,i And when i = j, g _i,j ＝1。

Each element G in the correlation matrix G _i,j The correlation between the channels of the user i and the user j is represented, and the correlation between the channels of the user i and the user j is calculated by an equation (5):

h _k ＝[h _k,1 ,h _k,2 ,...,h _k,n ,...h _k,N ] ^T (6)

wherein the base station is provided with N antennas, so that h is used _k,n Corresponding to the channel vector from the kth user to the nth antenna of the base station.

Let p be _k Represents the uplink transmit power, s, of the kth user _k Is the complex value data symbol sent by the kth user, z is the complex value vector of nx 1, representing additive white gaussian noise, the signal received by the base station is:

and 2, step: NOMA signal equalization

And (2) establishing an uplink transmission scene of the MISO-NOMA system based on the step 1, and carrying out equalization processing on the received signals by the base station.

Step 2.1: the base station performs equalization of signals by Maximum Ratio Combining (MRC), and the equalization matrix is expressed as:

W＝H ^* (8)

wherein H is a channel matrix between the base station and the user.

Under the condition of existing channel estimation, the base station utilizes the equalization method to realize the enhancement of a target signal and the weakening of an interference signal.

The signal after this Maximum Ratio Combining (MRC) equalization is expressed as:

wherein, the x and | | | | respectively represent the conjugate transpose and the vector norm; n is the number of antennas of the base station,

for the channel vector h _k The conjugate transpose of (1); y is a signal received by the base station;

is a k-th user signal weighted by a real-valued factor, where p _k Is the uplink transmit power, h, of the kth user _k A wireless channel vector between the base station and the k user; s _k Complex-valued data symbols sent for the kth user;

including intra-group interference and inter-group interference, k' is a subscript index representing users other than the k-th user;

is the noise component after equalization and z is additive white gaussian noise.

The maximum ratio combination can utilize the advantages of large-scale antennas, realize better balance performance under lower complexity, effectively reduce interference among users and improve the uplink transmission quality of user signals. By controlling the uplink transmit power of different users in a group, the interference between groups can be eliminated by the successive interference.

Step 2.2: in order to reduce the inter-group interference, it is necessary to optimize the user grouping method and reduce the correlation between the channel vectors of different groups of users. Suppose there is M in the first group _l A user, defining a vector pi _l ＝[π _l (1),π _l (2),…,π _l (M _l )]For the base station to the order of demodulation of the user signals of the l-th group, where _l (i) Subscript indicating the i-th demodulated user when the l-th group of user signals of the signal is demodulated by using the SIC.

Step 2.3: the base station firstly demodulates the subscript pi in the first group _l (1) To pi _l (i-1) and then removing these signal components from the received signal and demodulating the signal indexed by π _l (i) When the index in the l-th group is pi _l (i + 1) to π _l (M _l ) Is an interference component.

Use of

wherein, M _l' Is the number of users in the l' group;

channel vectors for the kth 'demodulated user in the l' user group;

And 3, step 3: NOMA user grouping method

In the uplink transmission process of users, it needs to determine which users reuse radio resources on the same time-frequency through grouping, i.e. belong to the same NOMA transmission group. Proper grouping can reduce inter-user interference and further save the total power consumption of the system.

The invention provides an uplink NOMA user grouping scheme based on K-means, which realizes effective grouping by utilizing the antenna correlation characteristic of a large-scale MIMO system and considering the channel difference characteristic among users in NOMA grouping.

The user grouping method comprises the following steps:

step 3.1: based on the optimal grouping number of SUS and the selection algorithm of the clustering center, the optimal grouping number L and the initial value corresponding to each group are calculatedClustering center C ⁰ 。

Step 3.2: and sequentially calculating the distance between each user to be grouped and each group of initial clustering centers (the calculation mode of the distance will be described in detail later), dividing the users into L groups, updating the clustering centers, and regrouping until the clustering centers are not changed any more.

The core of the above steps is the calculation of each group of corresponding initial clustering centers in step 3.1 and the calculation of each user to be grouped and each group of clustering centers C in step 3.2 ⁰ The distance of (2) is calculated.

The two core methods described above will be described next.

First, the present invention designs a method for identifying an optimal packet number L and an initial clustering center C using a Semi-Orthogonal User Selection (SUS) algorithm ⁰ The improvement of (1).

When the optimal number of packets L is designed, the too small number of packets may increase the demodulation interference signals associated with users in the group, which brings more challenges to the interference cancellation mechanism of the SIC. On the contrary, if the optimal number L of packets is too large, more interference will be generated between the transmission groups, which will affect the overall performance of the system. An appropriate optimal number L of packets should be as small as possible while at the same time distributing the least relevant users in different groups. If the initial clustering center can reflect this characteristic of the transport group, convergence of the grouping algorithm can be accelerated. The SUS algorithm can select users with better channel states and semi-orthogonal to each other by using the channel orthogonality degree between users, that is, the algorithm can calculate and find the least relevant user set in the group.

From this angle, an SUS algorithm is introduced to effectively find out the number of users that are least correlated in the group, that is, the optimal number of packets, and further determine an initial clustering center, thereby realizing effective grouping (capable of suppressing inter-group interference and reducing transmission power in subsequent power control).

The grouping characteristic of the NOMA transmission group and the channel characteristic of large-scale MIMO are fully utilized, the problem that K-means is sensitive to an initial clustering center is solved, and clustering convergence is accelerated.

step 3.1.2: initializing a candidate user set gamma, gamma = {1, \8230;, K }, wherein the candidate user set gamma represents all users which are not selected into the cluster center set, the cluster center set C is made to be an empty set, the optimal grouping number L is equal to 1, and the iteration index i is equal to 1.

Step 3.1.3: calculating a projection vector of each user k to a selected clustering center null space in a candidate user set gamma, wherein the projection vector of the user k to the selected clustering center null space is expressed as:

wherein h is _k Representing the radio channel vector between the base station and the k-th user.

wherein,

selecting subscript index corresponding to the clustering center;

step 3.1.5: updating a clustering center set S:

for the users which do not meet the semi-orthogonal requirement in the candidate user set gamma, the base station does not need to calculate the users in the subsequent iteration process, so the users can be directly abandoned in the updating step, and the complexity of the algorithm is reduced;

step 3.1.7: updating the optimal group number and the iteration index obtained by the current iteration round number, L ← L +1, i ← i +1;

step 3.1.9: finally, the initial clustering center C is obtained ⁰ ，

And an optimal number of packets L.

The algorithm effectively utilizes the diversity gain of multiple users, selects the clustering center with good channel quality and certain orthogonality, reduces the user interference among groups, and also improves the efficiency of subsequent user grouping.

Secondly, the invention provides a user grouping algorithm based on improved K-means, and the interference intensity after MRC equalization is reduced by reducing the channel correlation among different groups of users. The method specifically comprises the following steps:

step 3.2.1: according to the user channel matrix H = [ H = ₁ ,h ₂ ,…,h _K ] ^T Setting a weight factor epsilon and a balance factor lambda; the weighting factor epsilon is the weight of the total distance weight calculated for each distance term during the grouping process,the balance factor λ is a penalty value for balancing the group-to-user distance when there are too many users in a group.

Step 3.2.2: initializing cluster center set C ⁰ Make the cluster center set C ⁰ Is empty set and makes grouping matrix mu = mu ₁ ∪μ ₂ …∪μ _L T =0, the current iteration round number t is 0;

step 3.2.4: traversing each user in the candidate user set, and calculating the similarity of the user and each cluster center channel vector and the minimum channel gain difference of the user and each group;

more precisely, two channel vectors h are used _i And h _j By the cosine of (c) h _i And h _j The correlation of (c):

and

respectively normalizing the energy of the channel vectors corresponding to the ith and ith' users; omega _l Set of subscripts for users included in group I；

Step 3.2.5: on the basis of the formula (14) and the formula (15), a distance function is proposed for calculating the distance Dist from each user k to the l-th group in consideration of the group balance _k,l As shown in equation (16):

wherein ε ∈ (0, 1) represents the weight factor, when the total number of users in the l-th group M _l Above a certain threshold, e.g., 2K/L, a balancing factor λ is added when calculating the distance between the subsequent ungrouped users and the cluster center of the group, where λ > 1. The distance between the subsequent users and the group is increased, namely the possibility that the group is divided into more users is reduced, so that the number of users in different groups is balanced;

step 3.2.6: the group to which the user k belongs is selected,

wherein

Represents the index of the grouping closest to user k;

step 3.2.7: after the t-th round of clustering, the cluster center of each group is updated by formula (17):

Step 3.2.8: update the iteration index value, t ← t +1.

Step 3.2.9: repeating the steps 3.2.3-3.2.8, and updating the users in the group until the new C ^t No change occurs;

step 3.2.10: and outputting the user grouping matrix mu.

And 4, step 4: intra-group power control

After user grouping is completed, the invention provides a minimum uplink transmission power closed-form solution of users in a group considering minimum data rate constraint.

Step 4.1: taking an arbitrary group l in a cell as an example, assuming that there are M users in the group, a demodulation order pi is given _l ＝[π _l (1),π _l (2),…,π _l (M)]Assuming that the data rate constraints of all users in the ith group are equal and satisfy the minimum data rate constraint r:

wherein,

the achievable data rate for the ith demodulated user in the ith group;

to simplify the calculation, the target signal-to-interference-and-noise ratio (signal-to-interference-plus-noise ratio, SINR) requirement that the user needs to meet the minimum data rate constraint r is defined as γ:

wherein r is a minimum data rate constraint and W is a system bandwidth;

through analysis and derivation based on data rate constraints, the minimum uplink transmit power sum within the group can be calculated.

And 4.2: analysing the last demodulated user pi _l (M) the minimum data rate constraint to be satisfied is:

wherein,

are respectively the pi-th group _l The sum of uplink transmission power, channel vector, and intergroup interference plus noise of (M) demodulated users.

wherein, gamma is the target signal-to-interference-and-noise ratio requirement, N is the number of antennas of the base station, l, pi _l (M) is the index of the mth demodulation user in the first group after grouping,

respectively represent the l-th group pi _l (M) channel vectors of demodulated users and inter-group interference plus noise sums,

is a constant related to the target signal-to-interference-and-noise ratio requirement, the number of base station antennas and the user channels in the group.

Step 4.3: solving in the same way

And substituting the formula (22) to obtain the minimum uplink transmitting power of the M-1 user as follows:

step 4.4: the process is repeated continuously, and according to the calculation result, a general form of the minimum uplink transmission power required by any user i is derived and summarized as follows:

wherein,

channels of users corresponding to respective demodulation orders

And its conjugate transpose

According to the theorem, the invention obtains the power and closed-form solutions of all users in the same NOMA transmission group, realizes the balance of user power control in a cell in the integral power expression, and ensures the fairness among the users.

And 5: optimal user demodulation order within a group

The present invention uses SIC to demodulate users in the same NOMA transmission group. After the minimum uplink transmitting power in each group is obtained, the optimal demodulation sequence of the users in the group is given

Compared with the other demodulation sequences,

lower power consumption values can be achieved.

For an arbitrary NOMA transmission group, it is assumed that all users in the group have the same target data rate, such that

Indicating the user demodulation order. If and only if the user satisfies

When the user with higher interference demodulates first, the sum of the transmission powers of the group is the minimum, i.e. the demodulation order at that time

For an optimal demodulation order.

The optimality of this demodulation sequence can be demonstrated by a counter-proof method.

Step 6: inter-group power control

On the basis of the step 4 and the step 5, the invention provides an iterative power control method for calculating the inter-group power based on the strength of the inter-group interference, which solves the problem of coupling in the solving process caused by the inter-group interference and realizes the minimization of the total transmission power of the system under the condition of a determined NOMA transmission group.

First, in the classical power control problem, the quality of service requirement of uplink transmission users can be uniformly described by the following vector inequality:

p≥I(p) (25)

where I (p) is referred to as the standard interference function, which represents the interference that the user needs to overcome. P is the minimum transmit power that a user needs to meet a particular quality of service requirement.

Further, according to this constraint, the corresponding power minimization problem can be solved by the following iterative power control algorithm:

p ^(t+1) ＝I(p ^(t) ) (26)

where t represents the current iteration round.

For all minimum transmit powers p ≧ 0, when the above-mentioned interference function I (p) satisfies the following three properties:

1) Nonnegativity: i (p) is not less than 0

2) Monotonicity: if p is not less than p ', I (p) > I (p')

3) Expandability: for all δ ≧ 1, δ I (p) ≧ I (δ p)

I (p) is "standard" and its corresponding power minimization problem is a standard power control problem, which can be solved by the iterative power control algorithm in equation (26).

Furthermore, if the problem has an optimal solution, the iterative power control algorithm (26) will converge to the optimal solution.

Definition of

Wherein each element represents each set of corresponding interference; according to the step 4, the minimum uplink transmission power sum of users in the group is calculated

And the optimal demodulation order given in step 5

Each set of interference function I _l (q _-l ) The expression of (a) is:

wherein,

Channels of users corresponding to respective demodulation orders

And conjugate transpose thereof

And a constant related to the target SINR signal-to-interference-and-noise ratio gamma, k represents the index of the first j-1 demodulation orders, and k' is the order of the 1 st to the kth demodulation;

and

is the sum of the interference plus noise between the current ith group of demodulated users and the jth user demodulated thereafter, and the value of the sum is equal to the power q of other groups _-l And (4) correlating.

That is to say the interference function I _l (q _-l ) The minimum sum of transmit power of all users in the ith group to satisfy the minimum data rate constraint. And an interference function I is readily available _l (q _-l ) The three properties described above are satisfied, and therefore, we derive a "standard" interference function for each group based on the minimum transmit power sum of users in the optimal demodulation order within the group.

Furthermore, the power of each group can be solved through iteration, and the invention provides an iterative power control algorithm utilizing interference. According to the standard power control framework, given any non-null feasible power, the standard optimization problem in this region can converge to a unique optimal solution by an iterative method.

During each inter-group iteration, the transmit power sum of each target NOMA packet may be determined in turn:

wherein q is _l ^(t+1) Refers to the minimum sum of transmit power, M, of the l-th group in the t +1 th iteration _l Indicating the number of users of the group. The meaning of each component in the formula is the same as in formula (27).

The power of the other group l' ≠ l is fixed when calculating this group power. Determining power sum q of a group _l ^(t+1) Then, the power of each user in the group is calculated.

Under the given constraints of demodulation sequence and data rate, the minimum uplink transmission power of a user can be uniquely determined as follows:

wherein,

Through the analysis of the above power iteration strategy, the present invention summarizes the steps of the inter-group iteration power control algorithm in algorithm 3.

Step 6.2: initializing power sum vector q for each group ⁰ Power vector p for each user ⁰ The maximum iteration number maxter of the algorithm, and the current iteration round number t.

Step 6.3: in each iteration, sequentially traversing the users of each group to obtain the number M of users of the ith group _l And obtaining the optimal demodulation sequence obtained in step 5

Step 6.4: from the last demodulation, i.e. Mth _l The demodulation users start to calculate in sequence, and firstly obtain

And

wherein

In the optimal demodulation order

Next, index of current ith demodulation user in current ith group,

is the index of the j demodulation user that demodulates thereafter.

And

wherein,

represents the sum of the intergroup interference plus noise for the jth demodulated user demodulated after i.

Step 6.6: and then in the current iteration round number, calculating the minimum transmitting power sum q of the l group in the t +1 round iteration according to the formula (28) _l ^(t+1) ；

Step 6.7: sequentially calculating the minimum uplink transmitting power of each user i in the l group in the t +1 round iteration according to a formula (29)

Step 6.8: update the iteration index value, t ← t +1.

Step 6.10: outputting power sum vector of each group

And minimum uplink transmission power vector of each user

The algorithm can gradually converge to a unique fixed point for each group and an optimal power value through a finite number of iterations

And obtaining a global optimal power value. Within the feasible region, given an arbitrary power minimization problem instance with feasible solution, and an arbitrary initial power vector q ⁰ Then, according to the power iteration rule given in the formula (26), all the power values can be converged to the fixed unique optimal power value

And the most optimization of the total power of the system is realized.

In the following, the present application demonstrates the improvement of the system performance by the above techniques by comparing the impact of different power control schemes on the power consumption, energy efficiency and outage probability of the uplink MISO-NOMA system.

Wherein, the power consumption refers to the total transmission power of all users in the system, and is expressed as:

wherein q is _l Representing the sum of the transmit powers, mu, of the respective groups _l,k Grouping strategy, p, representing the ith group of k users _k And represents the uplink transmission power of the kth user.

The energy efficiency is calculated as follows:

wherein R is _k Indicates the achievable data rate, p, of the k-th user _k Representing the uplink transmission power of the kth user; the numerator represents the transmission data rate of all users in the system, and the denominator represents the total transmission power of all users in the system; the unit of energy efficiency is bits/J.

The system interrupt probability is calculated as the number of interrupt instances divided by the total number of simulation instances. And if the obtained user power consumption is larger than the maximum transmitting power, defining the user as an interrupt user. The presence of at least one interrupt user in the Monte Carlo simulation instance defines the instance as an interrupt instance.

The experimental results were obtained based on 5000 independently distributed channel instances randomly generated, with the cell radius set to 100 meters, and a total of K users evenly distributed within the cell. The system bandwidth W is 10MHz and the noise at the base station is set to-174 dBm/Hz. Fading channel models include small scale fading, distance dependent path loss and shadow fading. Specifically, small-scale fading follows a rayleigh distribution with a variance of 1; shadow fading follows a log normal model with a standard deviation of 10 dB; the path loss is 103.4+24.2log10 (d), where d is the Euclidean distance between the transmitting end and the base station.

In the simulation process, users in a cell are grouped through different schemes, and then the inter-group iterative power control algorithm provided by the invention is adopted to calculate the transmitting power based on a fixed clustering result. In the comparative experiment, the power control method proposed by the present invention was compared with a method combining user grouping and power control (hereinafter referred to as "Joint"), a Heuristic grouping method (hereinafter referred to as "heartbeat"), a Random grouping method (hereinafter referred to as "Random"), and a Non-grouped method (hereinafter referred to as "Non-clustered"), respectively.

In the Joint method, all the user grouping possibilities are traversed again in each iteration power control process to find the optimal grouping scheme until the total transmission power of the system converges to the minimum value. In the scheme, the global optimal solution can be found by the combined optimization of user grouping and power control. However, this method is extremely complicated and not practical in practice. In the Heuristic grouping method, users are firstly sequenced based on the channel gain, then the users are alternately distributed to each group based on the channel gain difference in the groups, and the sequence difference of the channel gain is formed in one NOMA group. Meanwhile, a random grouping scheme is also taken into consideration as a comparison algorithm. Meanwhile, in order to verify the effectiveness of the packet-based MISO-NOMA power control method, a comparison was also made with the Non-clustered power control method in which user grouping is not performed (each user is treated as an independent transmission group) in simulation experiments.

Fig. 4 compares the total power consumption and energy efficiency of the present invention with other methods in different user number scenarios, with the user data rate constraint set to 5Mbits/s. It can be seen that as the number of users increases, the power consumption difference between different methods increases gradually, and the method and the Joint method provided by the invention always maintain better and more stable performance than other methods when the number of users changes. In fig. 5, as the number of users increases, the energy efficiency difference between different methods gradually increases, and the method and the optimal Joint method according to the present invention achieve higher energy efficiency performance than other methods. At K =15, the proposed method can achieve 91%, 131%, and 225% energy efficiency improvements compared to the Heuristic, random, and Non-clustered methods. And achieves slightly lower performance than the method at a much lower complexity than the Joint scheme.

Fig. 6 shows the change of the outage probability under different data rate constraints for each method, and the number of users in a cell is set to 10. It can be seen that the duty cycle of the interrupt instance is almost zero when the data rate constraint is less than 5.5 Mbits/s. When the data rate constraint increases again, a user interruption begins to occur. When the data rate constraint reaches 7Mbits/s, 20% of users in the Non-shared method are in an interruption state, and the proportion of the number of interrupted users is 0.05% based on the method provided by the invention. With the increasing data rate constraint, the interruption probability of the method proposed by research and the Joint method is always lower than that of other methods.

Key words:

NOMA: non-Orthogonal Multiple Access

SIC: successive Interference Cancellation

MIMO: multiple-Input, multiple-Output, multiple-Input-Multiple-Output

MISO: multiple-Input, single-Output, multiple-Input-Single-Output

MRC: maximum Ratio combining, maximum Ratio combining

SUS: semi-Orthogonal User Selection

SINR: signal to noise ratio

QoS: quality of Service

Those not described in detail in this specification are well within the skill of the art.

Claims

1. A power control method of MISO-NOMA uplink channel is characterized by comprising the following steps:

step 2: the base station performs equalization processing on the received signal y through maximum ratio combining, optimizes a user grouping mode in order to reduce the interference among groups, reduces the correlation degree among channel vectors of different groups of users, and obtains

Represents the sum of the intergroup interference plus noise suffered by the ith demodulated user in the ith group;

and 4, step 4: obtaining a minimum uplink transmission power closed-form solution required by any user i in the group by considering minimum data rate constraint of the users in the group;

step 6: after the minimum uplink transmitting power and the optimal demodulation sequence required by any user i are obtained, the iterative power control method for calculating the inter-group power based on the strength of the inter-group interference realizes the minimization of the total transmitting power of the system under the condition of determining the NOMA transmission group.

2. The MISO-NOMA uplink channel power control method according to claim 1, comprising the steps of:

the step 1 specifically comprises the following steps:

step 1.1: establishing an uplink transmission scene of a single-cell MISO-NOMA system, wherein a base station simultaneously provides service for K users, K represents user serial number, K =1,2, \ 8230, K is provided with N antennas, N represents antenna serial number of the base station, N =1,2, \ 8230, N is provided with one antenna for each user, and assuming that all users in a cell are divided into L NOMA transmission groups, L represents the serial number of the NOMA transmission group, L =1,2, \\ 8230, L is less than N; defining a correlation matrix G with the size of K multiplied by K, which is used for representing the correlation of channel vectors between a base station and K users;

wherein,

establishing a channel matrix H between a base station and a user, H = [ H ] ₁ ,h ₂ ,…,h _K ] ^T The size is kxn, H is expressed as:

the channel vectors corresponding to different users are independent of each other, and each element V in the matrix V _k,n Can be decomposed into path loss beta _k,n And small scale fading ξ _k,n Where k and n are subscripts of the element in the matrix V, specifically V _k,n Expressed as:

due to the close distance between different base station antennas, the path loss beta experienced by the signal of the same user reaching different base station antennas can be considered _k,n Equal, small scale fading xi _k,n Are independently and identically distributed;

the correlation matrix G is a symmetric matrix satisfying:

wherein, for the element g of the ith row and the jth column in the matrix _i,j And element g of jth row and ith column _j,i Has g of _i,j ＝g _j,i And when i = j, g _i,j ＝1；

Each element G in the correlation matrix G _i,j Representing the correlation between the channels of the user i and the user j, wherein the correlation between the channels of the user i and the user j is calculated by an equation (5):

wherein (x) _i ,y _i ) And (x) _j ,y _j ) The representation being the position coordinates of two users, λ _c Denotes the correlation distance, λ _c The smaller the value of (a), the stronger the correlation between users in the cell;

h _k ＝[h _k,1 ,h _k,2 ,...,h _k,n ,...h _k,N ] ^T (6)

let p be _k Represents the uplink transmit power, s, of the kth user _k Is the complex-valued data symbol sent by the kth user, z is a complex-valued vector of nx 1, representing additive white gaussian noise, and the signal received by the base station is:

3. the method of power control for the MISO-NOMA uplink channel of claim 2, wherein step 2 comprises:

W＝H ^* (8)

wherein, H is a channel matrix between the base station and the user;

the signal after maximum ratio combining equalization is represented as:

wherein, ^* and | | | | represents conjugate transpose and vector norm respectively; n is the number of antennas of the base station,

is a channel vector h _k The conjugation transpose of (1); y is a signal received by the base station;

is a k-th user signal weighted by a real-valued factor, where p _k Uplink transmit power, h, for the kth user _k A wireless channel vector between the base station and the k user; s is _k Complex-valued data symbols sent for the kth user;

is the noise component after equalization, z is additive white gaussian noise;

step 2.2: in order to reduce the inter-group interference, it is necessary to optimize the user grouping method and reduce the correlation between the channel vectors of different groups of users, assuming that there is M in the l-th group _l Individual users, defining a vector pi _l ＝[π _l (1),π _l (2),…,π _l (M _l )]For the base station demodulation order of the user signals of the l-th group, where _l (i) Subscript representing the ith demodulated user when demodulating the ith group of user signals of the signal using SIC;

step 2.3: the base station firstly demodulates the subscript pi in the first group _l (1) To pi _l (i-1) and then removing these signals from the received signalDivide and then demodulate the subscript of pi _l (i) When the index in the l-th group is pi _l (i + 1) to π _l (M _l ) Is an interference component;

use of

wherein M is _l' The number of users in the l' th group;

the conjugate transpose of the channel vector of the ith demodulation user in the ith group;

channel vectors for the kth demodulated user within the l' user group; z is a radical of _l,πl(i) For the index in the ith user group of pi _l (i) The noise received by the user; and N is the number of antennas of the base station.

4. The method of power control for the MISO-NOMA uplink channel of claim 3, wherein step 3 comprises:

step 3.1: based on the optimal grouping number of SUS and the selection algorithm of the clustering center, the optimal grouping number L and the initial clustering center C corresponding to each group are calculated ⁰ ；

5. The method of power control for the MISO-NOMA uplink channel of claim 4, wherein step 3.1 comprises:

step 3.1.2: initializing a candidate user set gamma, gamma = {1, \8230;, K }, wherein the candidate user set gamma represents all users which are not selected into a cluster center set, the cluster center set C is made to be an empty set, the optimal grouping number L is equal to 1, and an iteration index i is equal to 1;

wherein,

selecting subscript index corresponding to the clustering center;

step 3.1.5:updating a clustering center set S:

step 3.1.6: updating the candidate user set according to the semi-orthogonality of the users in the candidate user set gamma and the clustering center set S to ensure that the users in the candidate user set are semi-orthogonal to the clustering center set, wherein the semi-orthogonality is determined by alpha:

step 3.1.7: updating the best group number and the iteration index obtained from the current iteration round number, L ← L +1, i ← i + 1';

step 3.1.9: finally, the initial clustering center C is obtained ⁰ ，

And an optimal number of packets L.

6. The MISO-NOMA uplink channel power control method of claim 5, wherein step 3.2 specifically comprises:

step 3.2.1: according to the user channel matrix H = [ H = ₁ ,h ₂ ,…,h _K ] ^T Setting a weight factor epsilon and a balance factor lambda; the weighting factor epsilon is a weighted value of each distance item in the total distance weight calculated in the grouping process, and the balance factor lambda is a punishment value for balancing the distance between a certain group and a user when the group of users is too many;

step 3.2.2: initializing cluster center set C ⁰ Make the cluster center set C ⁰ Is an empty set and makes the grouping matrix mu = mu ₁ ∪μ ₂ …∪μ _L T =0, whenThe number t of front iteration rounds is 0;

using two channel vectors h _i And h _j Is measured by the cosine of _i And h _j The correlation of (a):

the channel gain difference between users is measured using the squared difference of the normalized channel vectors, and the channel gain difference between user i and the l-th group is measured using the smallest gain difference within the group:

and

respectively normalizing the energy of the channel vectors corresponding to the ith and ith' users; omega _l Subscript sets for users included in the l-th group;

wherein ε ∈ (0, 1);

step 3.2.6: the group to which the user k belongs is selected,

wherein

Represents the index of the grouping closest to user k;

wherein M is _l The index of the users in the first group is represented by j;

step 3.2.8: updating an iteration index value, t ← t +1;

step 3.2.10: and outputting the user grouping matrix mu.

7. The MISO-NOMA uplink channel power control method of claim 6, wherein step 4 specifically comprises:

step 4.1: assuming that there are M users in a group, given a demodulation order π _l ＝[π _l (1),π _l (2),…,π _l (M)]Assuming that the data rate constraints of all users in the ith group are equal and the minimum data rate constraint r is satisfied:

wherein,

the achievable data rate for the ith demodulated user in the ith group;

defining the target signal-to-interference-and-noise ratio requirement which needs to be met by a user to meet the minimum data rate constraint r as gamma:

wherein r is a minimum data rate constraint and W is a system bandwidth;

calculating the minimum uplink transmitting power sum in the group through analysis and derivation based on the minimum data rate constraint r;

wherein,

are respectively the pi-th group _l (M) the sum of the uplink transmit power, channel vector, inter-group interference plus noise for the demodulation users;

Denotes the pi-th group in the l-th group _l The minimum uplink transmitting power of (M) demodulation users is obtained as follows:

wherein, gamma is the target signal-to-interference-and-noise ratio requirement, N is the number of antennas of the base station, and l, pi _l (M) is the index of the mth demodulation user of the ith group after grouping,

respectively represent the l-th group pi _l (M) channel vectors of demodulated users and the sum of inter-group interference plus noise,

is a constant related to the target signal to interference plus noise ratio requirement, the number of antennas of the base station and the user channels in the group;

step 4.3: solving in the same way

wherein,

channels of users corresponding to respective demodulation orders

And conjugate transpose thereof

8. The MISO-NOMA uplink channel power control method of claim 7, wherein step 5 specifically comprises:

after the minimum uplink transmitting power in each group is obtained, the optimal demodulation sequence of the users in the group is given

Indicating the user demodulation order, if and only if the user satisfies

When the user with higher interference demodulates first, the sum of the transmission power of the group is minimum, and the demodulation order is in this case

For an optimal demodulation order.

9. The method of power control for the MISO-NOMA uplink channel of claim 8, wherein step 6 comprises:

step 6.1: according to the user channel matrix H = [ H = ₁ ,h ₂ ,…,h _K ] ^T And a user number K, an antenna number N and a minimum data rate constraint r of the user;

step 6.2: initializing power sum vectors q for groups ⁰ Power vector p for each user ⁰ The maximum iteration number MaxIter of the algorithm and the current iteration round number t;

step 6.3: in each iteration, each iteration is traversed in turnThe group user obtains the number M of users in the first group _l And obtaining the optimal demodulation sequence obtained in step 5

Step 6.4: calculating in sequence from the first demodulation user to obtain

And

wherein

Is in the optimal demodulation order

Next, index of current ith demodulation user in current ith group,

is the index of the j demodulation user that demodulates thereafter;

step 6.5: calculating the values of all users in the current NOMA transmission group according to equation (27)

And

wherein,

representing the sum of the interclass interference plus noise for the current ith group of demodulated users,

inter-group interference representing the j demodulation user demodulating after iSum of disturbance plus noise;

wherein,

is a constant related to the target signal to interference and noise ratio requirement, the number of antennas of the base station and the user channel in the group;

channels of users corresponding to respective demodulation orders

And its conjugate transpose

and

is the sum of the interference plus noise between the current ith group of demodulated users and the jth user demodulated thereafter, and the value of the sum is equal to the power q of other groups _-l Correlation;

Wherein q is _l ^(t+1) Refers to the minimum sum of transmit power, M, of the l-th group in the t +1 th iteration _l Representing the number of users of the group;

Wherein,

step 6.8: updating an iteration index value, t ← t +1;

step 6.9: repeating the steps 6.3-6.8, and updating the power values of all groups and all users in the groups until t = MaxIter;

step 6.10: outputting power sum vector of each group

And minimum uplink transmission power vector of each user