Multi-user fair resource allocation method and system in MIMO-URLLC scene

Abstract

The invention provides a multi-user fair resource allocation method in an MIMO-URLLC scene. The method comprises the following steps: step 1, constructing a multi-user MIMO-URLLC communication system model; step 2, based on the multi-user MIMO-URLLC communication system model, constructing an original optimization problem; 3, equivalently converting the original optimization problem into a first optimization problem; step 4, by giving user throughput, converting the first optimization problem into a second optimization problem about power; step 5, splitting the second optimization problem into a plurality of sub-problems, and obtaining an optimal solution of each sub-problem; meanwhile, the value interval of the optimal user throughput is demonstrated, whether the power constraint is met or not under the condition of given user throughput is judged, if not, the user throughput is given again in the value interval of the optimal user throughput, and if yes, the final optimal solution is obtained. And if so, obtaining a final optimal solution. According to the method, a series of max min problems such as multi-user fair resource allocation are effectively solved, and a new thought is provided for solving the problems.

Description

Technical Field

[0001] The present invention relates to the field of communication technologies, and particularly to a method and system for multi-user fair resource allocation in a MIMO-URLLC scenario. Background Art

[0002] Ultra-Reliable Low Latency Communication (URLLC) is a core technology in modern wireless communication systems and plays a crucial role in 5G and future 6G systems. This technology aims to achieve communication latency in the millisecond or even sub-millisecond range and extremely high transmission reliability, mainly targeting application scenarios with extremely high requirements for latency and reliability such as industrial automation, remote healthcare, and unmanned driving. Multiple-Input Multiple-Output (MIMO) technology configures multiple antennas at the transmitter and receiver ends to achieve spatial multiplexing and diversity gain, thus significantly improving the transmission rate and system capacity of the wireless link. Multi-User Multiple-Input Multiple-Output (MU-MIMO) technology further expands on the basis of MIMO, allowing the base station to transmit data to multiple users simultaneously, effectively improving the spectrum utilization rate and the overall system throughput.

[0003] Although MU-MIMO brings higher efficiency to wireless communication in a multi-user environment, there are also significant challenges in resource allocation under the conditions of ensuring system performance and user fairness. Therefore, under the conditions of Finite Blocklength (FBL) and URLLC, for a certain type of MU-MIMO communication scenario, when ensuring multi-user fairness, a resource allocation search algorithm is designed for its achievable system performance and corresponding power allocation, effectively solving the communication resource allocation problem in this scenario. Summary of the Invention

[0004] To solve the problem of communication resource allocation among multiple users in an ultra-reliable low-latency scenario, the present invention proposes a multi-user fair resource allocation search method in a MIMO-URLLC scenario. The present invention adopts the following technical solutions:

[0005] In a first aspect, the present invention provides a multi-user fair resource allocation method in a MIMO-URLLC scenario, including:

[0006] Step 1. Construct a multi-user MIMO-URLLC communication system model;

[0007] Step 2. Based on the multi-user MIMO-URLLC communication system model, calculate the throughput of the single-user MIMO system, and construct the original optimization problem based on the throughput of the single-user MIMO system and the power constraint of the signal vector transmitted by the base station;

[0008] Step 3. Realize the separation of independent variables by expressing the single-user throughput as a function of SINR, and equivalently transform the original optimization problem into a first optimization problem;

[0009] Step 4. Transform the first optimization problem into a second optimization problem about power by given user throughput;

[0010] Step 5. Split the second optimization problem into several sub-problems, obtain the optimal solution of each sub-problem, so as to obtain the optimal solution of the second optimization problem; at the same time, demonstrate the value range of the optimal user throughput, and judge whether the power constraint is satisfied under the condition of the given user throughput in Step 4. If not, re-give the user throughput within the value range of the optimal user throughput, and return to Step 4. If satisfied, obtain the final optimal solution.

[0011] Further, the multi-user MIMO-URLLC communication system model includes:

[0012] Transmitter: A base station equipped with a total of n t transmit antennas;

[0013] Receiver: A total of m multi-antenna users: (u 1 , u 2 , …, u m ),

[0014] where the number of antennas corresponding to each user is (r 1 , r 2 , …, r m ), and the total number of antennas is n r , n r ≤ n t ;

[0015] The MIMO-URLLC communication system model supports URLLC services.

[0016] Further, the original optimization problem is:

[0017] max p min 1≤l≤m R l (p)

[0018] s.t. tr(P 2 Q * Q) ≤ p

[0019] where max is the maximum value, p is the transmission power of the base station, p is the power allocation vector, tr(P 2 Q * Q) is the signal vector transmitted by the base station, R l (p) is the throughput of the l-th multi-antenna user, Q is the normalized precoding matrix, Q * is the transpose of the normalized precoding matrix, and P is the power allocation matrix.

[0020] Furthermore, the throughput R l (p) of the l-th multi-antenna user is:

[0021]

[0022] where γ k is the SINR of the k-th antenna, k is the k-th antenna of user u l , V l represents the channel dispersion when the number of streams is r l , n is the block length of the FBL coding in the transmission process, ∈ is the maximum decoding error probability, and r l is the number of antennas corresponding to the l-th multi-antenna user.

[0023] Furthermore, the step 3 includes:

[0024] Denote tr(P 2 Q * Q) as:

[0025] Adopt ZF precoding, and make the throughputs of all users equal, and equivalently transform the original optimization problem into a first optimization problem;

[0026] where p 1 , p 2 , p i , are the powers allocated to the 1st, 2nd, i-th, and n-th r antennas respectively, and a 1 , a 2 , a i , are the power coefficients allocated to the 1st, 2nd, i-th, and n-th r antennas respectively.

[0027] Furthermore, the first optimization problem is:

[0028] max p,t t

[0029] s.t.t - R l (p) = 0, (1 ≤ l ≤ m)

[0030]

[0031] p i ≥σ 2 (1 ≤ i ≤ n r )

[0032] where σ 2 is the normalized noise power.

[0033] Furthermore, in step 4, for the determined user throughput t, the first optimization problem is transformed into a second optimization problem regarding power as follows:

[0034]

[0035] s.t. t - R l (p) = 0, (1 ≤ l ≤ m)

[0036] p i ≥σ 2 (1 ≤ i ≤ n r )

[0037] The optimal solution of the second optimization problem is pt * , verify whether it satisfies the power constraint; if not, decrease the value of t and reconstruct the second optimization problem; if satisfied, then p * = pt * , where p * is the optimal solution of the power allocation matrix, and the optimal solutions of the first optimization problem are t and pt * .

[0038] Furthermore, the splitting of the second optimization problem in step 5 includes: Denote the power allocation of user u 1 as The power allocation of user u l is The power allocation of user u 1 The corresponding power weight is The power weight of user u l is

[0039] For a certain determined t, construct m sub - problems, where the l - th sub - problem is:

[0040]

[0041] s.t. t - R l (p l ) = 0, (1 ≤ l ≤ m)

[0042] pu i ≥σ 2 (1 ≤ i ≤ r l)

[0043] Among them, r l is the number of receiving antennas of user u l ; are respectively the powers of the 1st, 2nd, …, rth l antennas of user u l ; are respectively the power weighting coefficients reflecting the channel states at the 1st, 2nd, …, rth l antennas of user u l ;

[0044] Furthermore, the value range for demonstrating the optimal user throughput in step 5 includes:

[0045] The power is evenly distributed, that is, each transmit power is allocated p eq as:

[0046]

[0047] At this time, the power pequ l after precoding for each user is:

[0048]

[0049] For the throughput of the user, m value range solving sub - problems are constructed. The lth value range solving sub - problem is:

[0050]

[0051] pu i ≥σ 2 (1 ≤ i ≤ r l )

[0052] The optimal solutions of the m value range solving sub - problems are respectively denoted as Req l (p l ),(1 ≤ l ≤ m)

[0053] Then the value range of the optimal solution is: [t min , t max , where

[0054] t max = max{Req 1 , Req 2 , …, Req m},

[0055] t min = min{Req 1 , Req 2 , …, Req m},

[0056] Req 1 , Req 2 , …, Req m Solve the optimal solutions of the sub-problems for the 1st, 2nd, …, mth value intervals respectively;

[0057] Take t max as the initial value t 0 for searching, and use the bisection method to search for t within the interval. When the interval length is less than the set threshold, take the minimum value of the interval as the approximation of t * .

[0058] On the other hand, the present invention provides a resource allocation system for multi-user fairness in the MIMO-URLLC scenario, including:

[0059] A communication system model construction module, which is used to construct a multi-user MIMO-URLLC communication system model;

[0060] A raw optimization problem construction module, which is used to calculate the throughput of a single-user MIMO system based on the multi-user MIMO-URLLC communication system model, and construct a raw optimization problem based on the throughput of the single-user MIMO system and the power constraint of the signal vector transmitted by the base station;

[0061] A first optimization problem transformation module, which is used to equivalently transform the raw optimization problem into a first optimization problem by separating the independent variables by expressing the single-user throughput as a function of SINR;

[0062] A second optimization problem transformation module, which is used to transform the first optimization problem into a second optimization problem about power by giving the user throughput;

[0063] An optimal solution acquisition module, which is used to split the second optimization problem into several sub-problems, obtain the optimal solutions of each sub-problem, so as to obtain the optimal solution of the second optimization problem; at the same time, demonstrate the value interval of the optimal user throughput, and judge whether the power constraint is satisfied under the condition of the given user throughput. If not, re-give the user throughput within the value interval of the optimal user throughput, and return to the previous module. If satisfied, obtain the final optimal solution.

[0064] Compared with the prior art, the present invention has the following beneficial effects:

[0065] 1. In view of a series of technical difficulties in the multi-user MMIO-URLLC communication scenario, such as the difficulty in constructing a model for the resource allocation problem considering multi-user fairness, numerous involved variables, and a complex solution process, under the FBL and URLLC conditions, for a certain type of MU-MIMO communication scenario, while ensuring multi-user fairness, a resource allocation search algorithm is designed for its achievable system performance and corresponding power allocation, effectively solving the communication resource allocation problem in this scenario. Different from the traditional iterative solution algorithm based on SCA (successive convex approximation), this algorithm does not need to perform convex approximation on the objective function. At the same time, through the reconstruction and dimensionality reduction of the problem, it has a lower complexity and can effectively solve a series of max min problems including but not limited to resource allocation with multi-user fairness, providing a new idea for solving such problems.

[0066] 2. The present invention proposes a resource allocation search algorithm for multi-user fairness in the MIMO-URLLC scenario. The core of the invention lies in the following four aspects: First, under the FBL and URLLC conditions, for a certain type of MU-MIMO communication scenario, while ensuring multi-user fairness, a resource allocation search algorithm is designed for its achievable system performance and corresponding power allocation, effectively solving the communication resource allocation problem in this scenario; Second, by proposing a lemma and giving the corresponding proof, the necessary conditions for a class of max min problems to obtain local optimal solutions are demonstrated in detail, providing inspiration for finding local optimal solutions to similar problems; Third, through the splitting and reconstruction of the problem, the dimensionality reduction of the optimization variables is achieved, thus greatly reducing the complexity of solving this problem; Fourth, different from the traditional iterative solution algorithm based on SCA (successive convex approximation), this algorithm does not need to perform convex approximation on the objective function and can effectively solve a series of max min problems including but not limited to resource allocation with multi-user fairness, providing a new idea for solving such problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0068] Figure 1 It is a flowchart of an embodiment of the present invention.

[0069] Figure 2 It is a schematic diagram for solving an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0070] To make the above objects, features, and advantages of the present application more apparent and understandable, the following provides a detailed description of the specific embodiments of the present application with reference to the accompanying drawings. A lot of specific details are set forth in the following description to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present application. Therefore, the present application is not limited by the specific embodiments disclosed below.

[0071] Embodiment 1

[0072] Combined with the attached Figure 1 drawings, the implementation flowchart of the optimization algorithm constructed by the present invention. The technical solution of the system in the following embodiments of the present invention is a multi-user fair resource allocation search method in a MIMO-URLLC scenario.

[0073] The model construction of the communication scenario includes the following: The transmitting end is a base station equipped with a total of t transmitting antennas; the receiving end has a total of m multi-antenna users (u 1 , u 2 , …, u m ), where the number of receiving antennas corresponding to each user is (r 1 , r 2 , …, r m ), and the total number of antennas is r; all users are considered through a single time-frequency resource block, and it is assumed that r ≤ t; the system supports URLLC services, the block length of the FBL coding during transmission is n, and the maximum decoding error probability is ∈; it is assumed that the antennas at the base station are uniform linear arrays, where the spatial correlation coefficient between adjacent antennas is ρ 0 ; denote the transmit power of the base station as p, and the power allocation matrix where p i represents the power allocated to the i-th of the total r antennas at the receiving end; denote the normalized precoding matrix as Then the signal vector Qx transmitted by the base station satisfies the corresponding power constraint.

[0074] Combined with the attached Figure 1 drawings, the implementation flowchart of the optimization algorithm constructed by the present invention. The specific technical solution of the system in the following embodiments is a multi-user fair resource allocation search algorithm in a MIMO-URLLC scenario, and the specific implementation steps are as follows:

[0075] Step 1. Construct a multi-user MIMO-URLLC communication system model;

[0076] In Step 1, the construction of the mathematical model of the multi-user MIMO communication system specifically includes: The transmitting end is a base station equipped with a total of n t transmitting antennas. The receiving end has a total of m multi-antenna users (u 1 , u2 , …, u m ), where the number of receiving antennas corresponding to each user is (r 1 , r 2 , …, r m ), and the total number of antennas is n r , that is

[0077]

[0078] In this embodiment, a single time-frequency resource block is used to consider the users, and it is assumed that n r ≤ n t . The system supports URLLC services. The block length of FBL coding during transmission is n, and the maximum decoding error probability is ∈. It is assumed that the antennas at the base station are a uniform linear array,

[0079] where the spatial correlation coefficient between adjacent antennas is ρ 0 , then the transmit correlation matrix

[0080]

[0081] where the spatial correlation coefficient between the i-th antenna and the j-th antenna The correlation matrix Denote the transmit power of the base station as p, and the power allocation matrix where p i represents the power allocated to the i-th of a total of n r antennas at the receiving end. Then the symbol vector transmitted at the transmit end can be expressed as

[0082] where The superscript T represents the transpose of the matrix.

[0083] Denote the normalized precoding matrix as Then the signal vector Qx transmitted by the base station satisfies

[0084] Power constraint

[0085]

[0086] where the superscript * represents the conjugate transpose of the matrix. Denote the normalized channel vector between the base station and the i-th receiving antenna Then there is

[0087]

[0088] where E t represents the n t -dimensional identity matrix. At this time, from the base station to the receiving end, there are a total of n rThe channel matrix of the root antenna can be expressed as

[0089]

[0090] If the regularized zero-forcing method is used for precoding, then we have

[0091]

[0092] where δ is a regulation coefficient, which is usually regarded as a constant. Considering complete elimination of interference between users in this model, the regulation coefficient δ in the above formula is set to 0. At this time, the regularized zero-forcing precoding degenerates into the conventional zero-forcing precoding. Thus, the signal vector at the receiving end can be given by the following formula

[0093] y = H T Qx + w (7)

[0094] where the additive white Gaussian noise and The channel vector in formula (4) above is normalized by the noise power, and here σ 2 is the normalized noise power.

[0095] Step 2. Based on the multi-user MIMO-URLLC communication system model, calculate the throughput of the single-user MIMO system, and construct an original optimization problem based on the throughput of the single-user MIMO system and the power constraint of the signal vector transmitted by the base station;

[0096] Based on the communication system model constructed in Step 1, deduce an optimization problem considering multi-user fairness and improving user throughput, specifically including: Denote the elements in the precoded channel matrix as At this time, the signal received by the k-th (1 ≤ k ≤ n r ) root antenna at the receiving end can be expressed as

[0097]

[0098] Denote the power allocation vector as The interference signal to the k-th root antenna is Then the SINR of the k-th root antenna can be expressed as

[0099]

[0100] Adjust the research object to the l-th (1 ≤ l ≤ m) multi-antenna user u l . For this user, this communication model is equivalent to a single-user MIMO system with t antennas at the transmitting end and r l antennas at the receiving end. Considering that r l ≤ n r ≤ nt , so it is equivalent that the number of streams of the transmitted signal is r l , and for the k-th (k ∈ u l ) antenna at the receiving end, its SINR is γ k . Then the throughput of this single-user MIMO system can be expressed as

[0101]

[0102] where ∈ represents the maximum decoding error probability during the transmission process, n represents the FBL coding block length, and Q -1 (·) represents a function inverse function of, and V l represents the channel dispersion when the number of streams is r l , that is

[0103]

[0104] Therefore, the original optimization problem is described as

[0105]

[0106] Step 3. Separate the independent variables by expressing the single-user throughput as a function of SINR, and equivalently transform the original optimization problem into a first optimization problem;

[0107] Demonstrate that in this optimization problem, under certain conditions, the optimal solution must be obtained when the throughputs of all users are equal. That is, when the SINRs of all users satisfy certain conditions (not less than 1) and the ZF precoding scheme is adopted, the conclusion that the optimal value is obtained when the throughputs of all users are equal holds. The specific demonstration is as follows:

[0108] First, prove a lemma.

[0109] Lemma 1: For an optimization problem in the form of

[0110] , denote its local optimal solution as (x * , y * , z * ). If it satisfies

[0111]

[0112] then h(x * , y * , z * ) = 0, and f(x * ) = g(y * , z * ).

[0113] Proof: The original optimization problem can be rewritten as an optimization problem in standard form.

[0114]

[0115] Denote the local optimal solution of this problem as (x * , y * , z * , t * ). According to the KKT conditions, we have

[0116]

[0117] Therefore, we have

[0118]

[0119] If λ 3 = 0, considering that f′(x * ) ≠ 0 and we can obtain λ 1 = λ 2 = 0, which contradicts λ 1 + λ 2 = 1. So λ 3 ≠ 0, that is, h(x * , y * , z * ) = 0. At this time, combined with equation (17), we know that λ 1 ≠ 0 and λ 2 ≠ 0, that is, f(x * ) = g(y * , z * ) = t * .

[0120] When there are more objective functions, as long as the independent variables of each objective function are independent of each other, it is easy to draw similar conclusions.

[0121] To express the original optimization problem (12) in a similar form, consider separating the independent variables by expressing the user throughput as a function of SINR. Let tr(P 2 Q * Q) be denoted as where a i ≤ 0, then we have

[0122]

[0123] At this time, the original optimization problem (12) can be expressed as

[0124]

[0125] Next, try to reconstruct its representation into the following form

[0126]

[0127] Among them, and R l The independent variables between (Γ) are independent of each other. The specific reconstruction process is as follows:

[0128] Considering that the SINR of each receiving antenna satisfies

[0129]

[0130] To solve this system of equations, we can first multiply the denominator polynomial to the left side of the equal sign, and then move all the terms on the left side to the right side of the equal sign, so that the original system of equations becomes a linear system of equations about p. Therefore, the algorithm complexity of solving this system of equations is It can be seen from Equation (21) that

[0131]

[0132] Among them, 1 ≤ i ≤ n represents a polynomial about (γ 1 , γ 2 , …, γ n ). Denote the polynomial inequality obtained by substituting into Equation (18) as

[0133]

[0134] Then there is

[0135]

[0136] So far, the reconstruction of the original optimization problem (12) has been completed. Combining Lemma 1, it is easy to obtain that for a specific local optimal solution, if the first-order information (derivative / partial derivative) of each objective function and constraint condition at this point is not zero, then it can be known that the function values of each objective function are equal when the optimum is achieved. The following is the proof: When the SINR satisfies certain conditions, the first-order information of the objective function is always not zero.

[0137] Lemma 2: When the SINR is not less than 1, that is, not less than 0 dB, the first-order (partial) derivative of the user throughput $R$ is always greater than 0.

[0138] Proof: When the number of user antennas is k, there is

[0139]

[0140] Generally, the value of ∈ is between 10 -5 and 10 -9 , and the block length n of the coding is generally dozens, hundreds or larger. If we denote

[0141]

[0142] Then it can be assumed that a < 1. Denote

[0143]

[0144] Then the function f(x 1 , x 2 , …, x k ) and R(x 1 , x 2 , …, x k ) have exactly the same monotonicity. It only needs to be proved that when x 1 ≥ 1, x 2 ≥ 1, …, x k ≥ 1 At this time, when x 1 ≥ 1, x 2 ≥ 1, …, x k ≥ 1, there is

[0145]

[0146] Similarly, it can be known that when x 1 ≥ 1, x 2 ≥ 1, …, x k ≥ 1 Therefore, when all SINRs

[0147] are not less than 1, all first-order partial derivatives of R are always greater than 0.

[0148] Furthermore, we consider the first-order information of the constraint condition g(γ 1 , γ 2 , …, γ n ). Different from the case of the objective function, this function does not have a unified form in different problems, which brings difficulties to analyzing the properties of its derivatives. However, in some cases, such as when the ZF precoding scheme is adopted, the interference between users is completely eliminated. At this time, the transformation between γ and p is a linear transformation, so that the constraint condition g is a linear constraint, and its first-order derivative is always non-zero. In summary, when all SINRs are not less than 1 and the ZF precoding is adopted, the optimal solution of this max min problem must be obtained when the throughputs of all users are equal.

[0149] According to the above conclusion, the original optimization problem in step 2 is equivalently transformed into the

[0150] first optimization problem, so that it can be solved by one-dimensional search. The specific process is as follows: Taking the ZF precoding scheme as an example. Combining the assumption that all SINRs of the above antennas are not less than 1, at this time p should satisfy

[0151] p i ≥ σ2 , (1 ≤ i ≤ r) (29)

[0152] The first optimization problem can be reconstructed into the following form:

[0153]

[0154] Step 4. By given user throughput, transform the first optimization problem into a second optimization problem regarding power;

[0155] To solve the above problem, a relatively large number t 0 (t 0 ≥ t * ) can be taken as the initial value for searching, and search from large to small. For a certain determined t, construct the following second optimization problem:

[0156]

[0157] Denote the optimal solution of this second optimization problem as p_t * , and verify whether it satisfies the power constraint. If not, decrease the value of t and reconstruct the sub - problem; if it satisfies, then p * = p_t * , and the optimal solution of the optimization problem (30) is jointly composed of the current t and p_t * .

[0158] Step 5. Split the said second optimization problem into several sub - problems, obtain the optimal solution of each sub - problem, thereby obtaining the optimal solution of the second optimization problem; meanwhile, demonstrate the value range of the optimal user throughput, and judge whether the power constraint is satisfied under the condition of the given user throughput in Step 4. If not, re - give the user throughput within the value range of the optimal user throughput and return to Step 4. If it satisfies, then obtain the final optimal solution.

[0159] Split the second optimization problem in Step 4 to achieve variable dimension reduction. The specific process is as follows: Denote the power allocation of user u 1 as The power allocation of user u l is Then there is

[0160]

[0161] And Denote the corresponding power weight of user u 1 as The power weight of user u l is Then there is

[0162]

[0163] At this time, there is R l (p) = R l (p l ) holds. For a certain determined t, m sub-problems are constructed, and the l-th sub-problem is:

[0164]

[0165] and Verify whether it satisfies the power constraint. If not, reduce the value of t and reconstruct the sub-problem; if it satisfies, then p * = pt * , and the optimal solution of the optimization problem (30) is jointly composed of the current t and pt * . Through the above method, the present invention splits an r-dimensional sub-problem into m sub-problems with lower dimensions, thereby realizing the simplification of the algorithm.

[0166] Set the value range of the optimal solution, and the specific steps are as follows: First, evenly distribute the power, that is, each transmit power is allocated as

[0167]

[0168] Denote the power after pre-coding of each user at this time as

[0169]

[0170] For the throughput of the user, construct m interval-solving sub-problems. The l-th interval-solving sub-problem is:

[0171]

[0172] Similar to the above sub-problem, this sub-problem can also be regarded as the existence problem of the intersection point of an n r -dimensional hyperplane and a high-dimensional surface. Therefore, the optimal solution must be obtained at the endpoints or tangent points of the feasible region.

[0173] So, as long as the objective function values at the endpoints of the feasible region and the tangent points of the surface are calculated, the optimal solution of this problem can be easily obtained. Denote the optimal solutions of these m sub-problems as

[0174] Req l (p l ),(1 ≤ l ≤ m) (38)

[0175] Let

[0176]

[0177] Then it is easy to know by the method of contradiction that there must be t min ≤ t* ≤ t max Therefore, t max can be used as the initial value t for the search 0 . Further, considering that finding the maximum value of t that satisfies the condition in the interval [t min , t max is a one-dimensional search problem, the process of the algorithm can be further optimized by some one-dimensional search methods. As Figure 2 shown, in this embodiment, the bisection method is used to search for t within the interval. When the interval length is less than the set threshold, the minimum value of the interval is taken as the approximation of t * , so that the optimal solution obtained by this method gradually approaches the theoretical optimal solution from below the theoretical optimal solution.

[0178] It should be noted that according to the needs of implementation, the various steps described in this application can be split into more steps, or two or more steps or partial operations of steps can be combined into new steps to achieve the purpose of the present invention.

[0179] Although this invention uses terms such as MIMO and SINR more frequently, it does not exclude the possibility of using other terms. Using these terms is only to more conveniently describe and explain the essence of this invention; interpreting them as any additional limitation is contrary to the spirit of this invention.

[0180] Embodiment 2

[0181] This embodiment provides a multi-user fair resource allocation system in a MIMO-URLLC scenario, including:

[0182] A communication system model construction module, which is used to construct a multi-user MIMO-URLLC communication system model; a raw optimization problem construction module, which is used to calculate the throughput of a single-user MIMO system based on the multi-user MIMO-URLLC communication system model, and construct a raw optimization problem based on the throughput of the single-user MIMO system and the power constraint of the signal vector transmitted by the base station;

[0183] A first optimization problem transformation module, which is used to equivalently transform the raw optimization problem into a first optimization problem by separating the independent variables by expressing the single-user throughput as a function of SINR;

[0184] A second optimization problem transformation module, which is used to transform the first optimization problem into a second optimization problem about power by giving the user throughput;

[0185] Optimal solution acquisition module, which is used to split the second optimization problem into several sub-problems, obtain the optimal solution of each sub-problem, so as to obtain the optimal solution of the second optimization problem; at the same time, demonstrate the value range of the optimal user throughput, and judge whether the power constraint is satisfied under the condition of the given user throughput. If not, re-give the user throughput within the value range of the optimal user throughput and return to the previous module. If satisfied, obtain the final optimal solution.

[0186] As described above, only the specific preferred embodiments of the present application are provided, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered by the protection scope of the present application.

[0187] It should be understood that the parts not detailed in this specification belong to the prior art. It should be understood that the above description of the preferred embodiment is relatively detailed, and it cannot be considered that this is a limitation on the protection scope of the present invention. Under the inspiration of the present invention, those of ordinary skill in the art can also make substitutions or deformations without departing from the scope protected by the claims of the present invention, and all fall within the protection scope of the present invention. The scope of protection requested by the present invention shall be subject to the appended claims.

Claims

1. A multi-user fair resource allocation method in a MIMO-URLLC scenario, characterized in that: include: Step 1. Construct a multi-user MIMO-URLLC communication system model; Step 2. Based on the multi-user MIMO-URLLC communication system model, the throughput of the single-user MIMO system is calculated, and the original optimization problem is constructed based on the throughput of the single-user MIMO system and the power constraint of the signal vector sent by the base station; Step 3. The original optimization problem is equivalently transformed into a first optimization problem by expressing the single-user throughput as a function of SINR to achieve separation of independent variables; Step 4. By giving the user throughput, the first optimization problem is transformed into a second optimization problem about power; Step 5. Split the second optimization problem into several sub-problems, obtain the optimal solution of each sub-problem, and thus obtain the optimal solution of the second optimization problem; at the same time, verify the value range of the optimal user throughput, and determine whether the power constraint is met under the condition of the given user throughput in step 4. If not, re-set the user throughput within the value range of the optimal user throughput, and return to step 4. If satisfied, obtain the final optimal solution.

2. The method for fair resource allocation for multiple users in a MIMO-URLLC scenario according to claim 1, characterized in that: The multi-user MIMO-URLLC communication system model includes: Transmitter: equipped with a total of n t Base station with a transmitting antenna; Receiver: A total of m multi-antenna users: (u1,u2,…,u m ), Among them, the number of antennas corresponding to each user is (r1, r2, …, r m ), and the total number of antennas is n r , n r ≤n t ; The MIMO-URLLC communication system model supports URLLC services.

3. The method for fair resource allocation for multiple users in a MIMO-URLLC scenario according to claim 2, characterized in that: The original optimization problem is: Among them, max is the maximum value, p is the transmission power of the base station, p is the power allocation vector, tr(P 2 Q * Q) is the signal vector sent by the base station, R l (p) is the throughput of the l-th multi-antenna user, Q is the normalized precoding matrix, Q * is the transpose of the normalized precoding matrix, and P is the power allocation matrix.

4. The method for fair resource allocation for multiple users in a MIMO-URLLC scenario according to claim 3, characterized in that: The throughput of the l-th multi-antenna user is R l (p) is: Among them, γ k is the SINR of the kth antenna, k is user u l The kth antenna, V l Indicates the number of flows is r l , n is the block length of FBL coding during transmission, ∈ is the maximum decoding error probability, r l is the number of antennas corresponding to the lth multi-antenna user.

5. The method for fair resource allocation for multiple users in a MIMO-URLLC scenario according to claim 4, characterized in that: The step 3 comprises: tr(P 2 Q * Q) is recorded as: Adopt ZF precoding and take the throughput of each user as equal, and transform the original optimization problem into the first optimization problem; Among them, p1, p2, p i , They are 1st, 2nd, i, nth r The power allocated to the antennas is a1, a2, a i , They are 1st, 2nd, i, nth r The power coefficient allocated to each antenna.

6. The method for fair resource allocation for multiple users in a MIMO-URLLC scenario according to claim 5, characterized in that: The first optimization problem is: Among them, σ 2 is the normalized noise power.

7. The method for fair resource allocation for multiple users in a MIMO-URLLC scenario according to claim 6, characterized in that: In step 4, for the determined user throughput t, the first optimization problem is transformed into a second optimization problem about power: The optimal solution of the second optimization problem is pt * , verify whether it meets the power constraint; if not, reduce the value of t and reconstruct the second optimization problem; if it meets the power constraint, then p * =pt * , where p * is the optimal solution of the power allocation matrix. The optimal solution of the first optimization problem is t and pt * .

8. The method for fair resource allocation for multiple users in a MIMO-URLLC scenario according to claim 7, characterized in that: The splitting of the second optimization problem in step 5 includes: recording the power allocation of user u1 as User l The power distribution is The corresponding power weight of user u1 is User l The power weight is For a certain t, construct m sub-problems, where the lth sub-problem is: Among them, r l For user u l The number of receiving antennas; For user u l 1st, 2nd, ..., r l The power of the antenna; Respectively reflect the user u l 1st, 2nd, ..., r l The power weighting coefficient of the channel state at the root antenna.

9. The method for fair resource allocation for multiple users in a MIMO-URLLC scenario according to claim 7, wherein: The value range for demonstrating the optimal user throughput in step 5 includes: The power is evenly distributed, that is, each transmission power is evenly distributed p eq for: At this time, the power of each user after precoding is pequ l for: According to the user throughput, m value intervals are constructed to solve sub-problems, among which the lth value interval solution sub-problem is: The optimal solutions to the subproblems in the m value intervals are denoted as Req l (p l ),(1≤l≤m) Then the optimal solution range is: [t min ,t max ],in t max =max{Req1,Req2,…,Req m }, t min =min{Req1,Req2,…,Req m }, Req1,Req2,…,Req m Find the optimal solution of the subproblem for the 1st, 2nd, ..., mth value intervals respectively; t max As the initial value t0 of the search, the binary search method is used to search for t in the interval. When the length of the interval is less than the set threshold, the minimum value of the interval is taken as the value of t. * Approximation of .

10. A multi-user fair resource allocation system in a MIMO-URLLC scenario, characterized in that: include: Communication system model building module. It is used to build a multi-user MIMO-URLLC communication system model; The original optimization problem construction module is used to calculate the throughput of the single-user MIMO system based on the multi-user MIMO-URLLC communication system model, and construct the original optimization problem based on the throughput of the single-user MIMO system and the power constraint of the signal vector sent by the base station; A first optimization problem conversion module, which is used to achieve separation of independent variables by expressing the single-user throughput as a function of SINR to convert the original optimization problem into a first optimization problem; The second optimization problem conversion module is used to convert the first optimization problem into a second optimization problem about power by giving a user throughput; Optimal solution acquisition module. It is used to split the second optimization problem into several sub-problems, obtain the optimal solution of each sub-problem, and thus obtain the optimal solution of the second optimization problem; at the same time, it verifies the value range of the optimal user throughput, and determines whether the power constraint is met under the condition of a given user throughput. If not, the user throughput is re-given within the value range of the optimal user throughput, and returns to the previous module. If it is met, the final optimal solution is obtained; the multi-user fair resource allocation system in the MIMO-URLLC scenario is used to execute the steps in the multi-user fair resource allocation method in the MIMO-URLLC scenario described in any one of claims 1-9.

Images