Power allocation method and apparatus for cf-ran, device, medium, and product

Abstract

The present application discloses a power allocation method and apparatus for a CF-RAN, a device, a medium, and a product. The method comprises: acquiring a channel estimation matrix between user equipments and access points, and beamforming matrices of edge distributed units; on the basis of the channel estimation matrix and the beamforming matrices, determining a signal-to-noise ratio for each user equipment using a pilot signal for downlink transmission, and determining spectral efficiency on the basis of the signal-to-noise ratio; by taking the maximization of the spectral efficiency as an objective, constructing a power allocation problem on the basis of an access point power constraint condition and a fronthaul capacity constraint condition; and solving the power allocation problem to obtain an optimal power allocation strategy. By providing the optimal power allocation strategy under the access point power constraint and the fronthaul capacity constraint, the problem that existing power allocation strategies cannot ensure the maximization of the spectral efficiency under a CF-RAN architecture is solved, improving the spectral efficiency under the fronthaul capacity constraint and the access point power constraint to the maximum extent, and improving system performance.

Description

Power allocation method, device, equipment, medium and product of CF-RAN

[0001] The present application claims priority to the Chinese patent application No. 202410444445.2 filed on April 12, 2024 with the China Patent Office, the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] The present application relates to the technical field of wireless communication, in particular to a power allocation method, device, equipment, medium and product of CF-RAN. BACKGROUND

[0003] Cell-free massive MIMO technology is used to solve the serious inter-cell interference occurring in ultra-dense networks and the high complexity of massive MIMO in traditional cellular networks.

[0004] For a distributed architecture, user merging can be implemented in different baseband units with the support of a pre-transmission network, thereby realizing scalability. However, since only a single access point (AP) is used for multi-user detection, the distributed architecture exhibits poor performance and obtains less gain compared to the implementation using small units. For a centralized architecture, cooperative processing between multiple APs can be implemented to improve network performance. However, when the number of cooperating APs is too large, the centralized processing faces the problems of high computational complexity and poor scalability.

[0005] A cell-free radio access network (CF-RAN) architecture combining the distributed and centralized methods is currently proposed to realize the scalability of cell-free. Under the CF-RAN architecture, the existing power allocation strategy cannot guarantee the maximum spectral efficiency.

[0006] SUMMARY

[0007] The present application provides a power allocation method, device, equipment, medium and product of CF-RAN to solve the problem that the existing power allocation strategy cannot guarantee the maximum spectral efficiency under the CF-RAN architecture, and gives a suitable power allocation method to maximize the spectral efficiency.

[0008] According to an aspect of the present application, a power allocation method of CF-RAN is provided, which is applied to a central processor of a cell-free radio access network (CF-RAN) system. The CF-RAN system further includes a user-centric distributed unit, an edge distributed unit and an access point configured on the edge distributed unit. The method includes:

[0009] obtaining a channel estimation matrix between a user equipment and the access point and a beamforming matrix of the edge distributed unit;

[0010] determine a signal-to-noise ratio of each user equipment using the pilot signal downlink transmission allocated to the user equipment according to the channel estimation matrix and the beamforming matrix, and determine a spectral efficiency according to the signal-to-noise ratio;

[0011] construct a power allocation problem according to an access point power constraint condition and a fronthaul capacity constraint condition, with the objective of maximizing the spectral efficiency;

[0012] solve the power allocation problem to obtain an optimal power allocation strategy.

[0013] According to another aspect of the present application, a power allocation device of a CF-RAN is provided, which is applied to a central processor of a CF-RAN system, the CF-RAN system further comprising: a user-centric distributed unit, an edge distributed unit, and an access point configured on the edge distributed unit; the device comprising:

[0014] an acquisition module, configured to acquire a channel estimation matrix between user equipment and the access point and a beamforming matrix of the edge distributed unit;

[0015] a spectral efficiency calculation module, configured to determine a signal-to-noise ratio of each user equipment using the pilot signal downlink transmission allocated to the user equipment according to the channel estimation matrix and the beamforming matrix, and determine a spectral efficiency according to the signal-to-noise ratio;

[0016] a problem construction module, configured to construct a power allocation problem according to an access point power constraint condition and a fronthaul capacity constraint condition, with the objective of maximizing the spectral efficiency;

[0017] a problem solving module, configured to solve the power allocation problem to obtain an optimal power allocation strategy.

[0018] According to another aspect of the present application, an electronic device is provided, comprising:

[0019] at least one processor; and

[0020] a memory connected with the at least one processor in communication; wherein,

[0021] the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the power allocation method of the CF-RAN according to any embodiment of the present application.

[0022] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for causing a processor to implement the power allocation method of the CF-RAN according to any of the embodiments of the present application when executed.

[0023] According to another aspect of the present application, a computer program product is provided, which comprises a computer program for implementing the power allocation method of the CF-RAN according to any of the embodiments of the present application when executed by a processor.

[0024] The technical solution of the embodiments of the present application obtains the channel estimation matrix between the user equipment and the access point, and the beamforming matrix of the edge distributed unit; determines the signal-to-noise ratio of the user equipment using the pilot signal downlink transmission according to the channel estimation matrix and the beamforming matrix, and determines the spectral efficiency according to the signal-to-noise ratio; constructs a power allocation problem according to the access point power constraint condition and the fronthaul capacity constraint condition with the goal of maximizing the spectral efficiency; solves the power allocation problem to obtain an optimal power allocation strategy. By giving the optimal power allocation strategy under the access point power constraint and the fronthaul capacity constraint, the problem that the existing power allocation strategy cannot guarantee the maximum spectral efficiency under the CF-RAN architecture is solved, the spectral efficiency under the fronthaul capacity constraint and the access point power constraint is maximized, and the system performance is improved.

[0025] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor is it used to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0026] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0027] Fig. 1 is a flowchart of a power allocation method of a CF-RAN according to an embodiment of the present application;

[0028] Fig. 2 is a structural schematic diagram of a CF-RAN system;

[0029] Fig. 3 is a flowchart of a power allocation method of a CF-RAN according to an embodiment of the present application;

[0030] Fig. 4 is a schematic diagram of simulation results and theoretical results of the number of pilots-spectral efficiency;

[0031] Figure 5 is a diagram of simulation results and theoretical results of the number of access points-spectrum efficiency;

[0032] Figure 6 is a diagram of simulation results and theoretical results of the number of antennas of an access point-spectrum efficiency;

[0033] Figure 7a is a diagram of convergence of an objective function of a PDD-based power allocation algorithm;

[0034] Figure 7b is a diagram of convergence of an objective function of a PDD-based power allocation algorithm under different numbers of EDUs;

[0035] Figure 8 is a diagram of the relationship between spectrum efficiency and fronthaul capacity;

[0036] Figure 9 is a diagram of the relationship between spectrum efficiency and the number of edge distributed units;

[0037] Figure 10 is a diagram of the structure of a power allocation device of a CF-RAN according to an embodiment of the present application;

[0038] Figure 11 is a diagram of the structure of an electronic device implementing a power allocation method of a CF-RAN according to an embodiment of the present application. DETAILED DESCRIPTION

[0039] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work should fall within the scope of protection of the present application.

[0040] It should be noted that the terms "first", "second", and "third" and the like in the description, claims, and drawings of the present application are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in other than the order illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to include only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to the process, method, product, or device.

[0041] Embodiment One

[0042] Figure 1 is a flow chart of a power allocation method of a CF-RAN according to an embodiment of the present application. The embodiment can be applied to a case where the power allocation of a CF-RAN system is implemented to maximize the spectral efficiency. The method can be performed by a power allocation device of the CF-RAN. The power allocation device of the CF-RAN can be implemented in the form of hardware and / or software, and can be configured in an electronic device.

[0043] Figure 2 is a structural diagram of a CF-RAN system. As shown in Figure 2, the CF-RAN system includes a central processing unit (CPU), a user-centric distributed unit (UCDU), an edge distributed unit (EDU), and an access point (AP) configured on the EDU. The UCDU mainly implements data distribution and combination; the EDU mainly implements channel estimation, multi-user / multi-stream detection, and multi-user / multi-stream beamforming functions; and the AP mainly serves as a radio frequency transceiver and performs digital-analog / analog-digital conversion.

[0044] In the CF-RAN system, one user equipment (UE) can be associated with multiple EDUs, but only one UCDU. In the uplink direction, the EDU estimates the uplink channel matrix between the AP and the user equipment, and selects the UE connected by the EDU according to the number of data streams supported by the system and the uplink demodulation reference signal. After the EDU detects the multi-user data stream, it is sent to the UCDU, which combines the same data stream from different user equipment (UE) EDUs. In the downlink direction, the beamforming matrix is calculated by the EDU, and the transmission data stream is beamformed.

[0045] As shown in Figure 1, the method includes:

[0046] S110, obtaining a channel estimation matrix between the user equipment and the access point, and a beamforming matrix of the edge distributed unit.

[0047] The channel estimation matrix is a matrix obtained by estimating the noise of the channel properties. The beamforming (BF) matrix can be calculated by the edge distributed unit (EDU) in the downlink direction. The goal of beamforming is to form the best combination or distribution of baseband (intermediate frequency) signals according to system performance indicators. Specifically, its main task is to compensate for signal fading and distortion introduced by spatial loss, multipath effects, and other factors during wireless transmission, while reducing interference between co-channel users.

[0048] In the embodiment, the CF-RAN system receives the signal transmitted by the user equipment UE through the receiving antenna of the access point AP of the edge distributed unit EDU, acquires the signal received by the edge distributed unit EDU through the central processing unit CPU, and obtains the channel estimation matrix through uplink channel estimation according to the signal between the user equipment UE and the access point AP. Moreover, in the downlink direction, the beamforming matrix is calculated by the edge distributed unit EDU, and the beamforming matrix is transmitted to the central processing unit, so that the central processing unit receives the beamforming matrix of the edge distributed unit EDU. The method for calculating the beamforming matrix can adopt the interference suppression based beamforming method, i.e. the zero-forcing algorithm (ZF). The main idea of the interference suppression based beamforming method is to find a beamforming vector under the condition that all real-time channel information of the user equipment is known, so that the interference of the target user equipment to all other user equipment is zero. The embodiment of the application will not be described here.

[0049] S120, determining the signal to noise ratio of the downlink transmission of the user equipment using the pilot signal according to the channel estimation matrix and the beamforming matrix, and determining the spectral efficiency according to the signal to noise ratio.

[0050] The signal to noise ratio, i.e. the signal to interference plus noise ratio (SINR), refers to the ratio of the strength of the received useful signal to the strength of the received interference signal (noise and interference).

[0051] In the embodiment, the signal to noise ratio of the downlink transmission of each user equipment UE using the pilot signal is calculated according to the channel estimation matrix and the beamforming matrix based on the definition of the signal to noise ratio and the large-dimensional random matrix theory, and the spectral efficiency is calculated according to the signal to noise ratio.

[0052] S230, constructing a power allocation problem according to the access point power constraint condition and the front haul capacity constraint condition with the maximum spectral efficiency as the target.

[0053] The access point power constraint condition refers to the condition required to be met by the power of each access point in the CF-RAN system. The front haul capacity constraint condition refers to the condition required to be met by the front haul capacity between the edge distributed unit EDU and the access point AP in the CF-RAN system. The front haul capacity is mainly determined by the wireless routing coding CPRI rate, and therefore, the front haul capacity constraint condition can also be represented by the rate between the edge distributed unit EDU and the access point AP.

[0054] In the embodiment, in order to obtain good system performance, the power allocation problem is constructed with the maximum spectral efficiency as the target and under the constraint of the access point power constraint condition and the front haul capacity constraint condition for the power allocation requirement.

[0055] S240, solving the power allocation problem to obtain an optimal power allocation strategy.

[0056] The optimal power allocation strategy is a strategy of allocating power to each user equipment when the spectrum efficiency reaches a maximum value, and can specifically include: allocating power to each user equipment.

[0057] In this embodiment, the constructed power allocation problem is solved by using a mathematical algorithm to obtain an optimal solution, so as to determine the optimal power allocation strategy when the spectrum efficiency reaches a maximum value.

[0058] The technical scheme of the embodiment of the application obtains a channel estimation matrix between user equipment and an access point, and a beamforming matrix of an edge distributed unit; determines a signal-to-noise ratio of downlink transmission of a pilot signal used by the user equipment according to the channel estimation matrix and the beamforming matrix, and determines spectrum efficiency according to the signal-to-noise ratio; constructs a power allocation problem according to an access point power constraint condition and a front-haul capacity constraint condition with the goal of maximizing the spectrum efficiency; and solves the power allocation problem to obtain an optimal power allocation strategy. By giving the optimal power allocation strategy under the access point power constraint and the front-haul capacity constraint, the power allocation problem is solved, the spectrum efficiency under the front-haul capacity constraint and the access point power constraint is maximized, and the system performance is improved.

[0059] Embodiment two

[0060] FIG. 3 is a flowchart of a power allocation method of a CF-RAN provided by the embodiment two of the application. The embodiment further refines the calculation method of the signal-to-noise ratio and the constructed power allocation problem on the basis of the above-mentioned embodiment. In this embodiment, the CF-RAN system includes Z edge distributed units EDU, N access points AP, and K user equipment UE. The N access points AP are allocated to the Z edge distributed units EDU, and each edge distributed unit EDU is connected to L access points AP. Each access point AP is equipped with M antennas, and each user equipment UE is equipped with one antenna. As shown in FIG. 3, the method includes:

[0061] S210, obtaining a channel estimation matrix between user equipment and an access point, and a beamforming matrix of an edge distributed unit.

[0062] In an optional embodiment, the channel estimation matrix between the K user equipment and the N access points connected to the Z edge distributed units is:

[0063] wherein, H z is an estimated channel matrix between the zth edge distributed unit and the K user equipment; is the estimated channel between the zth edge distributed unit and the kth user equipment, z ∈ [1, Z], k ∈ [1, K];

[0064] In the case that the qth user equipment using the tth pilot signal represents the kth user equipment, the estimated channel between the qth user equipment using the tth pilot signal and all the access points of the zth edge distributed unit is

[0065] The signal y received by the L access points of the zth edge distributed unit using the tth pilot signal through the receiving antennas is z,t

[0066] The channel h between the qth user equipment using the tth pilot signal and the zth edge distributed unit is z,t,q

[0067] Λ z,t,q = diag(λ z,t,q,1 ,…,λ z,t,q,L ) represents a large-scale channel fading matrix, λ z,t,q,L represents a large-scale fading factor of the qth user equipment using the tth pilot signal to the Lth access point of the zth edge distributed unit; g z,t,q ~ CN(0, I LM ) represents a small-scale channel fading matrix; n z,t ~ CN(0, εI LM ) represents a noise vector; I M is an M × M unit matrix, I N is an N × N unit matrix, I LM is an LM × LM unit matrix; ρ represents pilot signal power; τ represents the total number of pilot signals; ε represents noise power, S t represents a set of user equipments using the tth pilot signal.

[0068] In this embodiment, it is assumed that τ pilot signals are allocated to K UEs, and the channel between the qth UE using the tth pilot signal and all the EDUs is:

[0069] For uplink channel estimation, the signal received by the AP of the zth EDU using the tth pilot signal through the receiving antennas is:

[0070] ​​​According to a minimum mean square error (MMSE) channel estimation method, an estimated channel from the qth UE using the tth pilot signal to all EDUs is: wherein, and an estimated channel is obtained as: A channel estimation error is defined as A channel estimation error covariance matrix is A channel estimation matrix between L access points connected with all UEs (K user equipments) and all EDUs (Z edge distributed units) is: wherein is an estimated channel between the zth EDU and the kth UE; Λ t,q represents a large-scale fading factor from the qth user equipment using the tth pilot signal to all access points, n t represents a noise vector.

[0071] In another optional embodiment, a beamforming matrix of the zth edge distributed unit is:

[0072] wherein, W z is a beamforming matrix of the zth edge distributed unit; is an estimated channel matrix between the zth edge distributed unit and K user equipments; the upper index H is a matrix transposition symbol.

[0073] In this embodiment, for downlink channel transmission, it is assumed that all users are served by the zth EDU, and an estimated channel matrix between the zth EDU and K UEs is: A beamforming matrix of the zth EDU is:

[0074] S220, determining a signal-to-noise ratio of the kth user equipment using a pilot signal for downlink transmission according to the channel estimation matrix and the beamforming matrix.

[0075] In an optional embodiment, in the case that the qth user equipment using the tth pilot signal represents the kth user equipment, the signal-to-noise ratio of the kth user equipment is:

[0076] wherein, γ k is the signal-to-noise ratio of the kth user equipment, γ t,q is the signal-to-noise ratio of the qth user equipment using the tth pilot signal; γ k = γ t,q ; w z,i,j is a beamforming vector between the jth user equipment using the ith pilot signal and the zth edge distributed unit. is the beamforming vector between the qth user equipment using the tth pilot signal and the zth edge distributed unit; is the estimated channel between the qth user equipment using the tth pilot signal and all access points of the zth edge distributed unit;h z,t,q is the channel between the qth user equipment using the tth pilot signal and the zth edge distributed unit;P z = diag(p z,1 ,…,p z,K ) is the power coefficient matrix of the zth edge distributed unit, p z,k is the power coefficient of the zth edge distributed unit to the kth user equipment; p z,i,j denotes the power coefficient of the zth edge distributed unit to the jth user equipment using the ith pilot signal; p z,t,q denotes the power coefficient of the zth edge distributed unit to the qth user equipment using the tth pilot signal; is the downlink compressed noise of the zth edge distributed unit, q z,l ~ CN(0, μ z,l I M ) is the downlink compressed noise of the lth antenna of the zth edge distributed unit, μ z,l is the downlink compressed noise power, l ∈ [1, L]; is the noise variance of the qth user equipment using the tth pilot signal; the upper index H is the matrix transpose symbol.

[0077] wherein the power coefficient can be a precoding power allocation factor.

[0078] In the present embodiment, for downlink channel transmission, each AP needs to satisfy the power constraint, and the fronthaul link rate between EDUs is limited, and the signal processed by the zth EDU is x z = W z P z s+q z , wherein s = [s1,…, s K ] T = [s t,1 ,s t,2 ,…,s t,q ] T is the downlink transmission symbol of the K UEs (the qth UE using the tth pilot signal), and s k is a random variable with zero mean and zero unit variance, P z = diag(p z,1 ,…,p z,K ) is the power coefficient matrix of the zth EDU, is the downlink compressed noise of the zth EDU, q z,l ~CN(0,μ z,l I M is the downlink compressed noise of the zth EDU by the lth access point. The signal received by the qth UE using the tth pilot signal is:

[0079] wherein, is the additive white Gaussian noise of the qth UE using the tth pilot signal.

[0080] Definition is the beamforming vector between the zth EDU and the qth UE using the tth pilot signal, is the signal required by the qth UE using the tth pilot signal, s t,q denotes the signal sent to the qth user equipment using the tth pilot signal. In the case of the qth UE using the tth pilot signal representing the kth UE, the SINR of the qth UE using the tth pilot signal is:

[0081] wherein, is the power of the useful signal, is the interference term of other UEs to the kth UE, is the channel estimation error interference term, is the compressed noise interference term.

[0082] On the basis of the above-mentioned embodiments, in the case of calculating the beamforming matrix of the zth edge distributed unit using the zero-forcing algorithm, the spectral efficiency of the kth user equipment is:

[0083] R k = log2(1+γ k );

[0084] wherein, R k is the spectral efficiency of the kth user equipment, and γ k is the signal-to-noise ratio of the kth user equipment;

[0085] The asymptotic expression of the signal-to-noise ratio is:

[0086] wherein,

[0087] [Ξ z,t ] q,j is the matrix [Ξ z,tthe element in the qth row and jth column of matrix the element in the mth row and nth column of matrix denotes the channel estimation error interference matrix between all access points in the zth edge distributed unit for the qth user equipment using the tth pilot signal and the user equipment using the ith pilot signal; Λ z,t,q = diag(λ z,t,q,1 ,…,λ z,t,q,L ) denotes the large-scale channel fading matrix, λ z,t,q,L denotes the large-scale fading factor from the qth user equipment using the tth pilot signal to the Lth access point of the zth edge distributed unit; I N is an N x N identity matrix; ρ denotes the pilot signal power; τ denotes the total number of pilot signals; ε denotes the noise power, S t denotes the set of user equipment using the tth pilot signal; Λ z,k denotes the large-scale channel fading matrix corresponding to the channel between the kth user equipment and the zth edge distributed unit; denotes the noise variance of the kth user equipment; Tr(·) denotes the trace of a matrix.

[0088] In this embodiment, by using the theory of large-dimensional random matrix, the following can be obtained:

[0089] Let For the following can be obtained:

[0090] where a.s. means almost sure.

[0091] where:

[0092] Based on the properties of the zero-forcing algorithm precoding, the following can be obtained:

[0093] where (t, q)≠(i, j).

[0094] For the channel estimation error interference term in the denominator, it can be rewritten as:

[0095] the following can be obtained:

[0096] where:

[0097] For the compressed noise interference term, it can be rewritten as: ​

[0098] The spectral efficiency of the kth UE based on the zero-forcing algorithm beamforming is denoted as R k = log2(1 + γ k ), and the asymptotic form of γ k is denoted as:

[0099] When equal power allocation is used with the statistical power normalization factor,

[0100] The asymptotic form of γ k is rewritten as:

[0101] wherein, denotes the sum of the compressed noise interference and the Gaussian noise power; denotes the compressed noise interference, denotes the Gaussian noise.

[0102] S230, determining the spectral efficiency of the kth user equipment according to the signal-to-noise ratio of the kth user equipment.

[0103] In this embodiment, the spectral efficiency of the k user equipments is: R k = log2(1 + γ k ); wherein R k is the spectral efficiency of the kth user equipment, and γ k is the signal-to-noise ratio of the kth user equipment.

[0104] S240, constructing a power allocation problem according to the access point power constraint condition and the front-end capacity constraint condition with the maximum spectral efficiency as the target.

[0105] In an optional embodiment, the power allocation problem is expressed as:

[0106] s.t.

[0107] wherein,

[0108] det(·) denotes the determinant value of a matrix, Tr(·) denotes the trace of a matrix, log2(1 + γ k ) is the spectral efficiency of the kth user equipment, and γ kThe signal-to-noise ratio of the kth user equipment, k ∈ [1, K], K is the total number of user equipments; Λ is a large-scale channel fading matrix; I N is an N × N unit matrix; ρ is the pilot signal power; τ is the total number of pilot signals; ε is the noise power, S t is a set of user equipments using the tth pilot signal, μ z,l is the downlink compressed noise power.

[0109] In the embodiment, the single access point power constraint condition is specifically: the zth EDU will transmit the corresponding part of x z to its associated Aps, x z represents the received signal of the zth EDU. It is assumed that represents the lth AP received signal of the zth EDU: wherein, is the estimated channel matrix between the lth AP of the zth EDU and all UEs. The power constraint of the transmission signal of the lth AP of the zth EDU is: wherein, P max,z,l is the power limit of the lth AP of the zth EDU.

[0110] The fronthaul capacity constraint condition is specifically: represents the beamforming of the lth AP of the zth EDU. According to the rate-distortion theory, the rate between the zth EDU and the lth AP is represented as:

[0111] Therefore, the fronthaul capacity constraint is: wherein, C max,z,l is the fronthaul capacity between the zth EDU and the lth AP.

[0112] For the single AP power constraint condition, there is:

[0113] For the fronthaul capacity constraint, there is:

[0114] According to the large-dimensional random matrix theory, it can be obtained that:

[0115] wherein, represents almost sure convergence.

[0116] Therefore, it can be inferred that

[0117] S250, a first auxiliary variable {t k} is introduced, and the power allocation problem is equivalently represented as a first target problem.

[0118] In the present embodiment, the power allocation problem is a non-convex NP-hard problem, which is difficult to solve due to the strong coupling of multiple variables in the objective function of the power allocation problem. By using an efficient algorithm based on the PDD method, the problem is converted into a first target problem with a more easily handled form by introducing a first auxiliary variable k}.

[0119] In an alternative embodiment, the first target problem is expressed as:

[0120] s.t.

[0121] S260, introducing a second auxiliary variable , the first target problem is equivalently expressed as a second target problem.

[0122] In the present embodiment, in order to solve the coupling problem in the constraint , the first target problem is equivalently expressed as a second target problem by introducing a second auxiliary variable

[0123] In an alternative embodiment, the second target problem is expressed as:

[0124] s.t.

[0125] wherein,

[0126] S270, using a penalty dual decomposition method to solve the second target problem, obtaining the optimal power coefficient and the optimal compressed noise power.

[0127] In the present embodiment, in order to solve the coupling problem in the constraint ​The coupling problem in the second target problem is further solved by using a penalty dual decomposition (PDD) optimization method to decompose the second target problem into multiple sub-problems.

[0128] In an optional embodiment, the second target problem is solved by using a penalty dual decomposition method to obtain an optimal power coefficient and an optimal compression noise power, comprising:

[0129] A penalty coefficient η and a dual variable are introduced The second target problem is equivalently expressed as an augmented Lagrangian problem:

[0130] s.t.

[0131] wherein,

[0132] The augmented Lagrangian problem is decomposed into a first sub-problem, a second sub-problem and a third sub-problem;

[0133] The first sub-problem is expressed as:

[0134] s.t.

[0135] wherein, is a fixed value;

[0136]

[0137] The second sub-problem is expressed as:

[0138] s.t.

[0139] wherein, is a fixed value;

[0140] The third sub-problem is expressed as:

[0141] s.t.

[0142] wherein, is a fixed value;

[0143] In the n+1th calculation process, the first sub-problem is solved by using the concave-convex process algorithm to obtain a first optimal solution and a second optimal solution n≥0;

[0144] The second sub-problem and the third sub-problem are solved by using the Lagrange multiplier method to obtain a third optimal solution{p z,k ,μ z,l};

[0145] The maximum precision difference between the to-be-optimized variables and the second auxiliary variables is determined according to the first optimal solution, the second optimal solution and the third optimal solution; the to-be-optimized variables include the third optimal solution and the first auxiliary variables;

[0146] If the maximum precision difference is less than a preset precision parameter, the dual variables are updated according to a penalty coefficient as follows: Wherein, is the dual variable obtained in the n th calculation, is the dual variable obtained in the n+1th calculation;

[0147] If the maximum precision difference is greater than or equal to the preset precision parameter, the penalty coefficient is updated as follows:η n+1 =cη n ; c is a preset coefficient, and 0

[0148] The current calculation number n is increased by 1, and the n+1th calculation process is returned until the current calculation number n is greater than a maximum iteration number, or the maximum precision difference is less than a preset convergence precision, and the current third optimal solution is output, the current third optimal solution including an optimal power coefficient and an optimal compression noise power.

[0149] In this embodiment, all equality constraints and the penalty coefficient η are considered, and based on the PDD optimization method, the second target problem is augmented into an augmented Lagrange problem by using the dual variables as follows:

[0150] s.t.

[0151] The PDD algorithm has two levels of loops: the outer loop updates the dual variables and the penalty parameter, and the inner loop solves the above problem with fixed values of the dual variables and the penalty parameter.

[0152] In the inner loop, the optimization variables are partitioned into blocks, and the following steps are used to make the augmented Lagrangian problem decomposable into three subproblems that can be solved efficiently: a first subproblem Q1, a second subproblem Q2, and a third subproblem Q3.

[0153] First subproblem Q1: Fix and let The first subproblem can be formulated as:

[0154] s.t.

[0155] where,

[0156] There is a non-convex problem in the first subproblem, and the CCCP method is considered to solve it using the first Taylor expansion. Define the following:

[0157] and rewrite the constraints:

[0158] where, is a matrix with all elements equal to 1. Therefore, the first subproblem is rewritten as:

[0159] s.t.

[0160] The variables {t k}, and are decoupled and can be solved independently, where for {t k}, the following first subproblem with convexity is obtained:

[0161] For , the following first subproblem with convexity is obtained:

[0162] s.t.

[0163] For and , the CCCP algorithm based on PDD is used to solve, as shown in Algorithm 1:

[0164] Algorithm 2 CCCP algorithm based on PDD to solve S1

[0165] Input:

[0166] Initialize feasible point and

[0167] Set n = 0

[0168] 1 : repeat

[0169] 2 : solve first subproblem at current feasible point and find first and second optimal solutions and

[0170] 3 : set n = n + 1

[0171] 4 : update

[0172] 5 : until convergence

[0173] Output: first and second optimal solutions and

[0174] Second subproblem Q2: fix and Second subproblem is stated as:

[0175] s.t.

[0176] where Solved in closed form using Lagrange multiplier method.

[0177] S3 : fix and Subproblem is stated as

[0178] s.t.

[0179] where, Solved in closed form using Lagrange multiplier method.

[0180] For outer loop define the maximum constraint among all equality constraints: The equation holds when ω is small enough. In each outer loop iteration based on PDD algorithm, the dual variable is updated according to the constraint penalty parameter η and the dual variable. The dual variable is updated as follows: where, The PDD based power allocation algorithm is shown in Algorithm 2.

[0181] Algorithm 3: PDD-based power allocation algorithm

[0182] Input:

[0183] Define precision tolerance ω0, parameter Maximum number of iterations N O

[0184] Set n = 0

[0185] 1: Repeat

[0186] 2: Solve Q1 by CCCP algorithm

[0187] 3: Solve Q2 by Lagrange multiplier method

[0188] 4: Solve Q3 by Lagrange multiplier method

[0189] 5: if

[0190] 6: According to Update dual variables

[0191] 7: else

[0192] 8: η n+1 = cη n (0 < c < 1)

[0193] 9: end if

[0194] 10: Set n = n + 1

[0195] 11: Until ω < ω0or n > N O

[0196] Output: The third optimal solution {p z,k , μ z,l}.

[0197] The technical scheme of the embodiment of the application obtains a channel estimation matrix between a user equipment and an access point, and a beamforming matrix of an edge distributed unit; determines a signal-to-noise ratio of downlink transmission of a pilot signal used by the kth user equipment according to the channel estimation matrix and the beamforming matrix; determines the spectral efficiency of the kth user equipment according to the signal-to-noise ratio of the kth user equipment; constructs a power allocation problem according to the power constraint condition of the access point and the fronthaul capacity constraint condition with the maximum spectral efficiency as the target; introduces a first auxiliary variable {t k}, and equivalently represents the power allocation problem as a first target problem; introduces a second auxiliary variable The first target problem is equivalent to the second target problem; the second target problem is solved by using a penalty dual decomposition method to obtain optimal power coefficients and optimal compression noise power; the optimal power distribution strategy under the constraints of access point power and front-end capacity is given, and the power distribution problem is solved based on the highly non-convex optimization algorithm of the dual decomposition of the double ring penalty, which optimizes the power distribution coefficient and the compression noise power, and proposes a strategy combining distributed beamforming and centralized power distribution; the spectral efficiency under the constraints of front-end capacity and access point power is maximized, and the system performance is improved.

[0198] The power distribution method of the CF-RAN provided by the embodiment of the application calculates the power coefficient and the compression noise power in the CPU, and in the downlink direction of the CF-RAN system, each UCDU sends UE data to the EDU, each UCDU sends UE data to the EDU, and each EDU calculates a matrix Ξ z and Then, each EDU sends corresponding information based on the large-scale fading factor to the CPU. The CPU solves the power distribution problem and sends the power coefficient and the compression noise power to the EDU. Therefore, the EDU can calculate the ZF beamforming matrix according to the CSI and then send the signal to the AP. This algorithm realizes distributed beamforming and centralized power distribution with small cost in CPU and EDU data interaction.

[0199] Further, a plurality of performance comparison experiments are used to illustrate the advantages of the scheme of the application. In this embodiment, all access points AP are randomly associated with edge distributed units EDU, and each edge distributed unit EDU processes L APs. The channel model parameters are shown in Table 1.

[0200] Table 1

[0201] Figure 4 is a diagram of simulation results and theoretical results of pilot quantity-spectrum efficiency. Set M=8, N=48, K=10, the number of edge distributed units Z=1, Z=2, Z=4. Figure 4 shows that, in the case of random pilot strategy, as the pilot quantity increases, the spectrum efficiency also increases. The theoretical results are basically consistent with the simulation results, thereby proving the accuracy of the asymptotic expression of the signal-to-noise ratio. Figure 5 is a diagram of simulation results and theoretical results of access point quantity-spectrum efficiency. Set τ=4, K=6, M=8. Figure 5 shows that, as the AP quantity increases, the spectrum efficiency also increases. And because of the use of large-scale random matrix theory, the gap between the theoretical value and the simulation value decreases as the AP quantity increases. Figure 6 is a diagram of simulation results and theoretical results of access point antenna quantity-spectrum efficiency. Set N=48, K=6, τ=4. Z=1 represents the case of complete concentration, and Z=48 represents the case of complete distribution. Because the assumption of large-scale random matrix theory is not met in the case of complete distribution, only the simulation results of the antenna quantity-spectrum efficiency are given in Figure 6 when Z=48. It can be seen from Figure 6 that the performance of CF-RAN (Z=2, Z=4, Z=8) is between the case of complete concentration and the case of complete distribution, which is due to the better coordination ability of CF-RAN compared to the case of complete distribution. In addition, the spectrum efficiency of the system increases with the increase of the antenna quantity. Thus, the accuracy of the closed asymptotic spectrum efficiency expression is proved according to Figures 4 to 6.

[0202] Figure 7a is a diagram of the convergence of the objective function of the PDD-based power allocation algorithm; Figure 7b is a diagram of the convergence of the objective function of the PDD-based power allocation algorithm under different EDU quantities. Set N=20, M=8, K=6, τ=4, P max =1W, C max =4bps / Hz. In Figure 7a, the objective function value decreases with the increase of the iteration number, and converges after 15-20 iterations, with a convergence value of 52. In Figure 7b, the objective function value converges after 15 to 20 iterations. Moreover, the system performance when Z=2 is better than that when Z=1, because the complexity of the PDD-based power allocation algorithm is related to the number of edge distributed units Z. The more the number of EDUs, the less the cooperation within each EDU, and the more detailed the power allocation on the CPU. However, when the number of EDUs is too large, the mutual interference between them is more serious, resulting in performance degradation. As shown in Figure 7b, the system performance when Z=4 is worse than that when Z=1 and Z=2.

[0203] FIG. 8 is a schematic diagram of the relationship between spectral efficiency and fronthaul capacity. In the embodiment, M = 8, N = 20, K = 6, and τ = 4 are set. As shown in FIG. 8, when Z = 1, 2, and 4, the spectral efficiency increases with the increase of the fronthaul capacity, because the greater fronthaul capacity can reduce compression noise. When the fronthaul capacity Cmax> 16 bps / Hz, the spectral efficiency of the system converges under all the number of EDU embodiments, and the improvement of the spectral efficiency performance of the system is limited after the fronthaul capacity continues to increase. In addition, as shown in FIG. 7b, after multiple simulation averages, the performance of Z = 2 is better than that of Z = 1 and Z = 4, because the power allocation algorithm corresponding to Z = 2 is more complex than that of Z = 1, and the interference is smaller than that of Z = 4.

[0204] FIG. 9 is a schematic diagram of the relationship between spectral efficiency and the number of edge distributed units. In order to ensure the effectiveness of the large-scale random matrix theory, N = 24, M = 16, K = 6, and τ = 4 are set, and the spectral efficiency under four conditions of different power constraints of each AP is compared. The spectral efficiency of the system increases with the increase of the power limit of the AP. With the increase of the number of EDUs, the spectral efficiency performance of the CF-RAN system presents a trend of first rising and then falling. This is because when the number of EDUs is small, with the increase of the complexity of the power allocation algorithm, the system performance also improves, and when the number of EDUs is greater than 2, the internal interference of the system becomes the main factor affecting the system performance. Therefore, according to FIGS. 7 to 9, the superiority of the power allocation algorithm based on PDD is proved.

[0205] Embodiment Three

[0206] FIG. 10 is a schematic diagram of the structure of a power allocation device of a CF-RAN provided in Embodiment Three of the present application. As shown in FIG. 10, the device comprises an acquisition module 310, a spectral efficiency calculation module 320, a problem construction module 330, and a problem solving module 340.

[0207] The acquisition module 310 is configured to acquire a channel estimation matrix between user equipment and the access point and a beamforming matrix of the edge distributed unit.

[0208] The spectral efficiency calculation module 320 is configured to determine the signal-to-noise ratio of each user equipment using the allocated pilot signal for downlink transmission according to the channel estimation matrix and the beamforming matrix, and determine the spectral efficiency according to the signal-to-noise ratio.

[0209] The problem construction module 330 is configured to construct a power allocation problem according to the access point power constraint condition and the fronthaul capacity constraint condition, with the maximum spectral efficiency as the target.

[0210] The problem solving module 340 is configured to solve the power allocation problem to obtain an optimal power allocation strategy.

[0211] Optionally, the CF-RAN system comprises: Z edge distributed units, N access points and K user equipments; the N access points are distributed to the Z edge distributed units, each edge distributed unit is connected with L access points; each access point is equipped with M antennas, and each user equipment is equipped with 1 antenna.

[0212] Optionally, the beamforming matrix of the zth edge distributed unit is:

[0213] wherein W z is the beamforming matrix of the zth edge distributed unit; is the estimated channel matrix between the zth edge distributed unit and the K user equipments; the upper index H is the matrix transposition symbol.

[0214] Optionally, the channel estimation matrix between the K user equipments and the N access points connected with the Z edge distributed units is:

[0215] wherein, is the estimated channel matrix between the zth edge distributed unit and the K user equipments; is the estimated channel between the zth edge distributed unit and the kth user equipment, z∈[1,Z], k∈[1,K];

[0216] In the case that the qth user equipment using the tth pilot signal represents the kth user equipment, the estimated channel between the qth user equipment using the tth pilot signal and all the access points of the zth edge distributed unit is :

[0217] The signal y z,t received by the L access points of the zth edge distributed unit using the tth pilot signal through the receiving antennas is:

[0218] The channel h z,t,q between the qth user equipment using the tth pilot signal and the zth edge distributed unit is:

[0219] Λ z,t,q =diag(λ z,t,q,1 ,…,λ z,t,q,L ) represents the large-scale channel fading matrix, λ z,t,q,L represents the large-scale fading factor from the qth user equipment using the tth pilot signal to the Lth access point of the zth edge distributed unit; g z,t,q~CN(0, I LM ) denotes a small-scale channel fading matrix; n z,t ~CN(0, εI LM ) denotes a noise vector; I M is an M x M identity matrix, I N is an N x N identity matrix, I LM is an LM x LM identity matrix; p denotes a pilot signal power; t denotes a total number of pilot signals; ε denotes a noise power, S t denotes a set of user equipments using the t-th pilot signal.

[0220] Optionally, in the case that the q-th user equipment using the t-th pilot signal represents the k-th user equipment, a signal-to-noise ratio of the k-th user equipment is:

[0221] wherein γ k denotes a signal-to-noise ratio of the k-th user equipment, γ t,q denotes a signal-to-noise ratio of the q-th user equipment using the t-th pilot signal; γ k = γ t,q ; W z is a beamforming matrix of the z-th edge distributed unit; w z,i,j denotes a beamforming vector between the j-th user equipment using the i-th pilot signal and the z-th edge distributed unit; denotes a beamforming vector between the q-th user equipment using the t-th pilot signal and the z-th edge distributed unit; denotes an estimated channel between the q-th user equipment using the t-th pilot signal and all access points of the z-th edge distributed unit; h z,t,q denotes a channel between the q-th user equipment using the t-th pilot signal and the z-th edge distributed unit; P z = diag(p z,1 ,…,p z,K ) denotes a power coefficient matrix of the z-th edge distributed unit, p z,k denotes a power coefficient of the z-th edge distributed unit to the k-th user equipment; p z,i,j denotes a power coefficient of the z-th edge distributed unit to the j-th user equipment using the i-th pilot signal; p z,t,q denotes a power coefficient of the z-th edge distributed unit to the q-th user equipment using the t-th pilot signal; denotes a downlink compressed noise of the z-th edge distributed unit, q z,l ~CN(0, μ z,l I Mdenotes the downlink compressed noise of the lth access point at the zth edge distributed unit, μ z,l denotes the downlink compressed noise power, l ∈ [1, L]; denotes the noise variance of the qth user equipment using the tth pilot signal; the upper index H is the matrix transpose symbol.

[0222] Optionally, in the case of calculating the beamforming matrix of the zth edge distributed unit by using the zero-forcing algorithm, the spectral efficiency of the kth user equipment is:

[0223] R k = log2(1 + γ k );

[0224] wherein R k is the spectral efficiency of the kth user equipment, and γ k is the signal-to-noise ratio of the kth user equipment;

[0225] The asymptotic expression of the signal-to-noise ratio is:

[0226] wherein,

[0227] [Ξ z,t ] q,j is the element of the qth row and the jth column of the matrix [Ξ z,t ]; is the element of the mth row and the nth column in the matrix ; denotes the channel estimation error interference matrix between all access points in the zth edge distributed unit of the qth user equipment using the tth pilot signal and the user equipment using the ith pilot signal; Λ z,t,q = diag(λ z,t,q,1 ,…,λ z,t,q,L ) denotes the large-scale channel fading matrix, λ z,t,q,L denotes the large-scale fading factor of the qth user equipment using the tth pilot signal to the Lth access point of the zth edge distributed unit; I N is an N × N unit matrix; ρ denotes the pilot signal power; τ denotes the total number of pilot signals; ε denotes the noise power, S t denotes the set of user equipment using the tth pilot signal; Λ z,k denotes the large-scale channel fading matrix corresponding to the channel between the kth user equipment and the zth edge distributed unit; denotes the noise variance of the kth user equipment; Tr(·) denotes the trace of a matrix.

[0228] Optionally, the power allocation problem is expressed as:

[0229] s.t.

[0230] wherein,

[0231] det(·) denotes the determinant of a matrix, Tr(·) denotes the trace of a matrix, log2(1+γ k ) is the spectral efficiency of the kth user equipment, γ k is the signal-to-noise ratio of the kth user equipment, k∈[1,K], K is the total number of user equipments; Λ is a large-scale channel fading matrix; I N is an N×N identity matrix; ρ is the pilot signal power; τ is the total number of pilot signals; ε is the noise power, S t is the set of user equipments using the tth pilot signal.

[0232] Optionally, the problem solving module 340 comprises:

[0233] a first equivalent unit, configured to introduce a first auxiliary variable {t k}, and equivalently express the power allocation problem as a first target problem; the first target problem is expressed as:

[0234] s.t.

[0235] introducing a second auxiliary variable , and equivalently express the first target problem as a second target problem, the second target problem is expressed as:

[0236] s.t.

[0237] wherein,

[0238] Solving the second target problem by using a penalized dual decomposition method to obtain an optimal power coefficient and an optimal compression noise power.

[0239] Optionally, the problem solving unit is specifically configured to:

[0240] introducing a penalty coefficient η and a dual variable equivalently representing the second target problem as an augmented Lagrangian problem:

[0241] s.t.

[0242] wherein,

[0243] decomposing the augmented Lagrangian problem into a first sub-problem, a second sub-problem and a third sub-problem;

[0244] in the n+1th calculation process, solving the first sub-problem by using a concave-convex process algorithm to obtain a first optimal solution and a second optimal solution n≥0;

[0245] solving the second sub-problem and the third sub-problem by using a Lagrange multiplier method to obtain a third optimal solution z,k ,μ z,l};

[0246] determining a maximum precision difference between the to-be-optimized variables and the second auxiliary variables according to the first optimal solution, the second optimal solution and the third optimal solution; the to-be-optimized variables include the third optimal solution and the first auxiliary variables;

[0247] if the maximum precision difference is less than a preset precision parameter, updating the dual variables according to a penalty coefficient: wherein, is the dual variable obtained in the n th calculation, is the dual variable obtained in the n+1th calculation;

[0248] If the maximum precision difference is greater than or equal to a preset precision parameter, the penalty coefficient is updated as: η n+1 =cη n wherein c is a preset coefficient, 0 < c < 1.

[0249] The current calculation number n is added by 1, and the (n+1)th calculation process is returned until the current calculation number n is greater than the maximum iteration number, or the maximum precision difference is less than the preset convergence precision, and the current third optimal solution is output, the current third optimal solution comprising: an optimal power coefficient and an optimal compressed noise power.

[0250] Optionally, the first sub-problem is expressed as:

[0251] s.t.

[0252] wherein, a and b are fixed values.

[0253] The second sub-problem is expressed as:

[0254] s.t.

[0255] wherein, a and b are fixed values.

[0256] The third sub-problem is expressed as:

[0257] s.t.

[0258] wherein, a and b are fixed values.

[0259] The power allocation device of the CF-RAN provided in the embodiments of the present application can execute the power allocation method of the CF-RAN provided in any of the embodiments of the present application, and has the corresponding function modules and beneficial effects of executing the method.

[0260] Embodiment Four

[0261] ​​​FIG. 11 shows a structural diagram of an electronic device 10 that can be used to implement embodiments of the present application. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular telephones, smartphones, wearable devices (e.g., headsets, glasses, watches, etc.), and other similar computing devices. The components shown here, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present application as described and / or claimed in this document.

[0262] As shown in FIG. 11, the electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., connected to the at least one processor 11 in communication, where the memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer programs stored in the read-only memory (ROM) 12 or loaded into the random access memory (RAM) 13 from the storage unit 18. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0263] Various components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc., an output unit 17, such as various types of displays, speakers, etc., a storage unit 18, such as a magnetic disk, an optical disk, etc., and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0264] The processor 11 can be various general and / or special purpose processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 performs various methods and processes described above, such as the power allocation method of the CF-RAN.

[0265] In some embodiments, the power allocation method of the CF-RAN can be implemented as a computer program tangibly embodied in a computer readable storage medium, e.g., storage unit 18. In some embodiments, portions of the computer program or all of the computer program can be loaded onto the electronic device 10 via, e.g., ROM 12 and / or communication unit 19. When the computer program is loaded onto RAM 13 and executed by processor 11, one or more steps of the power allocation method of the CF-RAN described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the power allocation method of the CF-RAN by any other suitable means, e.g., by means of firmware.

[0266] Various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, specially designed application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0267] In some embodiments, the power allocation method of the CF-RAN can be implemented as a computer program tangibly embodied in a computer program product, the computer program implementing the power allocation method of the CF-RAN of the present application when executed by a processor, the computer program product can be understood as a software product that mainly realizes the solution of the present application through the computer program. The computer program for implementing the method of the present application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the computer program, when executed by the processor, implements the functions / operations specified in the flow charts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as a separate software package and partially on a remote machine, or entirely on a remote machine or server.

[0268] In the context of this application, a computer readable storage medium can be a tangible medium that can contain or store computer programs for use by or in connection with an instruction execution system, apparatus, or device. Computer readable storage media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer readable storage medium can be a machine readable signal medium. More specific examples of the machine readable storage medium will include a one or more lines of a electrical connection, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0269] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0270] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), blockchain network, and the Internet.

[0271] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0272] It should be understood that the steps shown above in various forms of flow can be reordered, added, or deleted. For example, each step described in the present application can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions of the present application can be achieved, and the present application is not limited herein.

[0273] The above detailed description does not constitute a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A CF-RAN power allocation method, applied to a central processor of a cellular-free radio access network (CF-RAN) system, wherein the CF-RAN system further comprises: A user-centric distributed unit, an edge distributed unit, and an access point configured on the edge distributed unit; the method includes: Acquire a channel estimation matrix between the user equipment and the access point and a beamforming matrix of the edge distributed unit; Determining a signal-to-noise ratio (SNR) of downlink transmission of the allocated pilot signal by each user equipment based on the channel estimation matrix and the beamforming matrix, and determining a spectrum efficiency based on the SNR; With the goal of maximizing the spectrum efficiency, a power allocation problem is formulated according to access point power constraints and fronthaul capacity constraints; The power allocation problem is solved to obtain an optimal power allocation strategy.

2. The method according to claim 1, wherein The CF-RAN system includes: Z edge distributed units, N access points and K user devices; N access points are allocated to the Z edge distributed units, each edge distributed unit is connected to L access points; each access point is equipped with M antennas, and each user device is equipped with 1 antenna.

3. The method according to claim 2, wherein: The beamforming matrix of the zth edge distributed unit is: Among them, W z represents the beamforming matrix of the zth edge distributed unit; represents the estimated channel matrix between the z-th edge distributed unit and K user equipments; the superscript H is the matrix transpose symbol.

4. The method according to claim 2, wherein: The channel estimation matrix between K user devices and N access points connected to Z edge distributed units is: in, represents the estimated channel matrix between the zth edge distributed unit and K user equipments; represents the estimated channel between the zth edge distributed unit and the kth user equipment, z∈[1,Z],k∈[1,K]; In the case where the qth user equipment using the tth pilot signal represents the kth user equipment, the estimated channels between the qth user equipment using the tth pilot signal and all access points of the zth edge distributed unit are for: The signal y received by the L access points of the zth edge distributed unit using the tth pilot signal through the receiving antenna z,t for: The channel h between the qth user equipment and the zth edge distributed unit using the tth pilot signal z,t,q for: Λ z,t,q =diag(λ z,t,q,1 ,…,λ z,t,q,L ) represents the large-scale channel fading matrix, λ z,t,q,L represents the large-scale fading factor from the qth user equipment to the Lth access point of the zth edge distributed unit using the tth pilot signal; g z,t,q ~CN(0,I LM ) represents the small-scale channel fading matrix; n z,t ~CN(0,εI LM ) represents the noise vector; I M is the M×M identity matrix, I N is the N×N identity matrix, I LM is the LM×LM unit matrix; ρ represents the pilot signal power; τ represents the total number of pilot signals; ε represents the noise power, S t represents the set of user equipments using the tth pilot signal.

5. The method according to claim 4, wherein When the qth user equipment using the tth pilot signal represents the kth user equipment, the signal-to-noise ratio of the kth user equipment is: Among them, γ k represents the signal-to-noise ratio of the kth user equipment, γ t,q represents the signal-to-noise ratio of the qth user equipment of the tth pilot signal; γ k =γ t,q ;W z is the beamforming matrix of the zth edge distributed unit; w z,i,j represents the beamforming vector between the jth user equipment and the zth edge distributed unit using the i-th pilot signal; is the beamforming vector between the qth user equipment using the tth pilot signal and the zth edge distributed unit; represents the estimated channel between the qth user equipment and all access points of the zth edge distributed unit using the tth pilot signal; h z,t,q represents the channel between the qth user equipment and the zth edge distributed unit using the tth pilot signal; P z =diag(p z,1 ,…,p z,K )express The power coefficient matrix of the zth edge distributed unit, p z,k represents the power coefficient of the zth edge distributed unit to the kth user equipment; p z,i,j represents the power coefficient of the zth edge distributed unit to the jth user equipment using the i-th pilot signal; p z,t,q represents the power coefficient of the zth edge distributed unit to the qth user equipment using the tth pilot signal; represents the downlink compression noise of the zth edge distributed unit, q z,l ~CN(0,μ z,l I M ) represents the downlink compression noise of the lth access point at the zth edge distributed unit, μ z,l represents the downlink compression noise power, l∈[1,L]; represents the noise variance of the qth user equipment using the tth pilot signal; the superscript H is the matrix transpose symbol.

6. The method according to claim 5, wherein: When the zero-forcing algorithm is used to calculate the beamforming matrix of the z-th edge distributed unit, the spectrum efficiency of the k-th user equipment is: R k =log2(1+γ k ); Among them, R k is the spectrum efficiency of the kth user equipment, γ k is the signal-to-noise ratio of the kth user equipment; The asymptotic expression of the signal-to-noise ratio is: in, [Ξ z,t ] q,j is the matrix [Ξ z,t ]’s element in the qth row and jth column; is a matrix The element at row m and column n in ; represents the channel estimation error interference matrix between the qth user equipment using the tth pilot signal and the user equipment using the ith pilot signal and all access points in the zth edge distributed unit; Λ z,t,q =diag(λ z,t,q,1 ,…,λ z,t,q,L ) represents the large-scale channel fading matrix, λ z,t,q,L I represents the large-scale fading factor from the qth user equipment to the Lth access point of the zth edge distributed unit using the tth pilot signal; N is an N×N unit matrix; ρ represents the pilot signal power; τ represents the total number of pilot signals; ε represents the noise power, S t represents the set of user equipment using the t-th pilot signal; z,k represents the large-scale channel fading matrix corresponding to the channel between the k-th user equipment and the z-th edge distributed unit; represents the noise variance of the kth user equipment; Tr(·) represents the trace of the matrix.

7. The method according to claim 6, wherein: The power allocation problem is expressed as: in, det(·) represents the determinant of the matrix, Tr(·) represents the trace of the matrix, log2(1+γ k ) represents the spectrum efficiency of the kth user equipment, γ k represents the signal-to-noise ratio of the kth user equipment, k∈[1,K], K represents the total number of user equipment; Λ represents the large-scale channel fading matrix; I N is an N×N unit matrix; ρ represents the pilot signal power; τ represents the total number of pilot signals; ε represents the noise power, S t represents the set of user equipments using the tth pilot signal.

8. The method according to claim 7, wherein: Solving the power allocation problem to obtain an optimal power allocation strategy includes: Introduce the first auxiliary variable {t k }, the power allocation problem is equivalently expressed as the first objective problem; the first objective problem is expressed as: Introducing the second auxiliary variable The first objective problem is equivalently expressed as the second objective problem, which is expressed as: in, The penalty dual decomposition method is adopted to solve the second objective problem and obtain the optimal power coefficient and the optimal compressed noise power.

9. The method according to claim 8, wherein The second objective problem is solved by using a penalty dual decomposition method to obtain an optimal power coefficient and an optimal compressed noise power, including: Introducing penalty coefficient η and dual variable The second objective problem is equivalently expressed as an augmented Lagrangian problem: in, Decomposing the augmented Lagrangian problem into a first subproblem, a second subproblem, and a third subproblem; In the n+1th calculation process, the concave-convex process algorithm is used to solve the first subproblem and obtain the first optimal solution. and the second optimal solution n≥0； The Lagrange multiplier method is used to solve the second and third subproblems, and the third optimal solution {p z,k ,μ z,l }; Determine the maximum precision difference between the variable to be optimized and the second auxiliary variable according to the first optimal solution, the second optimal solution, and the third optimal solution; the variable to be optimized includes: the third optimal solution and the first auxiliary variable; If the maximum precision difference is less than the preset precision parameter, the dual variable is updated according to the penalty coefficient as follows: in, is the dual variable obtained by the nth calculation, is the dual variable obtained by the n+1th calculation; If the maximum precision difference is greater than or equal to the preset precision parameter, the penalty coefficient is updated to: η n+1 =cη n ; c is the preset coefficient, 0 <c<1; Add 1 to the current number of calculations n and return to the n+1th calculation process until the current number of calculations n If the maximum number of iterations is greater than the maximum precision difference, or the maximum precision difference is less than the preset convergence precision, the current third optimal solution is output, and the current third optimal solution includes: an optimal power coefficient and an optimal compressed noise power.

10. The method according to claim 9, wherein: The first sub-problem is stated as: in, is a fixed value; The second sub-problem is expressed as: in, is a fixed value; The third sub-problem is expressed as: in, is a fixed value; 11. A CF-RAN power distribution device, applied to a central processor of a cellular-free radio access network (CF-RAN) system, wherein the CF-RAN system further comprises: A user-centric distributed unit, an edge distributed unit, and an access point configured on the edge distributed unit, the device comprising: An acquisition module, configured to acquire a channel estimation matrix between the user equipment and the access point and a beamforming matrix of the edge distributed unit; a spectrum efficiency calculation module, which determines a signal-to-noise ratio (SNR) of downlink transmission of the allocated pilot signal by each user equipment based on the channel estimation matrix and the beamforming matrix, and determines a spectrum efficiency based on the SNR; A problem construction module, which constructs a power allocation problem based on access point power constraints and fronthaul capacity constraints with the goal of maximizing the spectrum efficiency; The problem solving module solves the power allocation problem and obtains the optimal power allocation strategy.

12. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, where the computer program is executed by the at least one processor to enable the at least one processor to perform the power allocation method for the CF-RAN according to any one of claims 1 to 10.

13. A computer-readable storage medium storing computer instructions, wherein the computer instructions are configured to cause a processor to implement the CF-RAN power allocation method according to any one of claims 1 to 10 when executed.

14. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the computer program implements the power allocation method of the CF-RAN according to any one of claims 1 to 10.

Images