A network slice resource allocation method with deterministic quality of service guarantee

Abstract

The application provides a network slice resource allocation method with deterministic quality of service guarantee, comprising the following steps: (1) constructing a multi-slice user-centered cell-free massive multiple input multiple output (CF-mMIMO) network architecture; (2) based on the quality of service requirements of each user equipment, constructing a resource allocation optimization problem with the maximum sum rate of all users as the target; (3) eliminating the non-convexity of the short packet rate formula through a successive rate lower bound maximization algorithm; (4) eliminating the non-convexity of the signal-to-noise ratio formula through a successive convex approximation (SCA) algorithm; (5) optimizing the bandwidth resources and power resources through an alternating optimization method; (6) optimizing the user slice association variables through a simulated annealing algorithm; and (7) calculating the maximum sum rate according to the optimal allocation result. The application significantly improves the sum rate, reduces the transmission delay, improves the resource utilization efficiency, and has good expansibility and practical value.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of 6G wireless communication, specifically a network slice resource allocation method with deterministic quality of service guarantee, which is suitable for intelligent power grids, autonomous driving and other scenarios that require ultra-low latency and high reliability. BACKGROUND

[0002] With the rapid development of wireless communication technology, the sixth generation of mobile communication technology (6G) has become a hot spot for research and development. 6G networks need to support ultra-low latency, ultra-high reliability and large-scale access performance indicators to meet the complex needs of intelligent manufacturing, autonomous driving, telemedicine and other vertical industries. Network slicing technology is a virtualized resource management technology that creates multiple logically isolated virtual networks (slices) on the same physical network to achieve customized support for different business needs. 6G has an urgent need for network slicing, because traditional "one-size-fits-all" network design cannot meet the differentiated needs of businesses in high-dynamic, complex scenarios. Through network slicing, 6G can effectively overcome low physical network resource utilization, inadequate QoS guarantees, and other issues, providing isolation and deterministic quality of service guarantees for different business scenarios.

[0003] Smart grids, as a deep integration of power and information communication technology, are an important support for the transformation of the power industry to efficient, safe, stable and intelligent. With the increasing demand for communication networks by smart grids, the limitations of the fifth generation of public network technology (5G) are gradually emerging. Although 5G provides ultra-high bandwidth and low latency connectivity, it still cannot fully meet the extremely high reliability and strict latency requirements of smart grids in control-type businesses. For example, power grid protection, load control and other critical tasks require near "fiber-level" performance, including millisecond-level latency and near 100% data transmission reliability. Moreover, 5G's network slicing technology still has shortcomings in resource isolation and hard quality guarantees, especially in complex scenarios where multiple slices are shared, making it difficult to avoid resource competition and uncertainty. This gap makes it difficult for 5G to meet the hard expectations of smart grids for control-type business communication quality.

[0004] The emergence of 6G provides an important opportunity to solve the above problems. 6G can flexibly allocate network resources according to the diversified needs of smart grids through more refined network slice management and optimized architecture design. Especially in control-type business scenarios, 6G can provide stronger resource exclusivity and guarantee ultra-reliable and low-latency communication services through customized high-priority slices. In addition, 6G introduces sub-6GHz, millimeter wave, and terahertz frequency bands, etc. in frequency band selection, which significantly improves data transmission rate and network capacity, providing hardware foundation support for low-latency and high-reliability control-type businesses. However, in smart grid communication networks, resource allocation based on network slicing still faces many challenges. Due to the diversity of smart grid businesses, the QoS requirements of different businesses differ significantly, and existing resource allocation methods are difficult to develop efficient resource allocation strategies based on these differences. In addition, in the unique network architecture of 6G, how to reasonably allocate frequency and power resources based on network slicing technology through access selection and resource optimization mechanisms to meet the stringent requirements of smart grid control-type businesses for reliability and low latency is still a hot issue that needs to be studied.

[0005] In summary, the existing 5G network slicing technology has deficiencies in resource isolation and hard QoS guarantee, and it is difficult to meet the millisecond-level latency and ultra-high reliability requirements of smart grid control-type businesses. Traditional resource allocation methods do not consider the short packet rate characteristics of URLLC businesses, and resource competition is easy to occur in multi-slice sharing scenarios. The present application solves the above problems by combining CF-mMIMO architecture and new optimization algorithms. SUMMARY

[0006] Technical problem: The present application provides a network slicing resource allocation method with deterministic quality of service guarantee to solve the problems of unreasonable resource allocation leading to resource waste, not applying the actual 6G network scenario architecture, and not considering the QoS constraints of URLLC businesses in the prior art. This method maximizes the sum rate of all users while meeting the deterministic QoS requirements of each user, realizes the reasonable allocation of system resources, improves the sum rate, reduces the transmission delay, and avoids resource waste.

[0007] Technical solution: The technical problem to be solved by the present application is to provide a network slicing resource allocation method with deterministic quality of service guarantee, which maximizes the sum rate of all users as the optimization target based on guaranteeing the QoS requirements of users, and solves the optimization problem through successive rate lower bound maximization algorithm, SCA algorithm, and simulated annealing algorithm to realize the reasonable allocation of resources.

[0008] To solve the above technical problems, the technical scheme adopted by the present application is:

[0009] The application discloses a network slice resource allocation method with deterministic quality of service guarantee, which comprises the following steps: constructing a multi-slice user-centered cell-free massive multiple input multiple output (CF-mMIMO) network architecture, and defining the connection relationship between an access point (AP) and a user equipment (UE); based on the QoS demand of each UE, constructing a resource allocation optimization problem with the maximum sum rate of all users as the target;

[0010] decomposing a complex non-convex problem; eliminating the non-convexity of a short packet rate formula through a successive rate lower bound maximization algorithm; eliminating the non-convexity of a signal-to-noise ratio formula through an SCA algorithm; alternately optimizing bandwidth resource allocation and power resource allocation to obtain optimal results; obtaining optimal user slice association variables through a simulated annealing algorithm; and calculating the maximum sum rate according to the optimal resource allocation results.

[0011] Preferably, the CF-mMIMO network distributes a large number of APs and UEs, the AP set is M={1,..., m,..., M}, the UE set is U={1,..., i,..., U}, all the APs are connected with a central processing unit (CPU), bandwidth resources are divided into physical resource blocks (PRBs) with equal bandwidth, the total number of PRBs is K, the UEs are allocated to different slices, the slice set is S={1,..., s,..., S}, each PRB can be used by all the UEs in one slice, all the APs transmit data to the connected UEs, interference is generated between the UEs in the same slice, the UEs and the APs communicate through ultra-reliable low latency communication (URLLC) data packets, and the UE i can tolerate an upper limit of a transmission delay t i,b .

[0012] Preferably, in the CF-mMIMO network, the UE selects the AP set serving the UE by measuring a large-scale channel gain, selects an AP satisfying a channel gain not less than a preset threshold, and selects an AP with the largest gain if there is no such AP. The construction rule of the service set of the UE is:

[0013] C i ={m|α i,m ≥β i,m (d0)}∪{argmax m∈M α i,m}

[0014] where C i denotes the set of APs serving UE i, α i,m , β i,m denote the large-scale fading parameter and path loss between UE i and AP m, respectively, β i,m is a function of distance d, and d0is a pre-defined distance threshold for path loss. Each AP determines the connection relationship with UEs based on the serving set of UEs using 0-1 variable c i,m , which denotes the connection relationship between UE i and AP m, c i,m = 1 if there is a connection, otherwise c i,m = 0.

[0015] Preferably, the optimization variables of the constructed resource allocation optimization problem are: the association variable ω i,s of UE i with slice s, the number of PRBs allocated to slice s δ s , and the power allocation factor ξ i,m associated with UE i-AP m. The signal-to-interference-plus-noise ratio Γ i at UE i on a single PRB is defined as:

[0016]

[0017] where ρ0and P denote the total transmit power of a single AP and the number of antennas, respectively, γ i,m denotes the independent channel gain on a single antenna between UE i and AP m. Based on the formula of Γ i , the achievable short packet data rate of UE i performing URLLC service on a single PRB is:

[0018]

[0019] where W denotes the bandwidth of a single PRB, τ c and τ denote the period of coherence interval and the length of pilot sequence, respectively, n denotes the length of URLLC data packet block, Q -1 (·) denotes the inverse function of Gaussian Q function, ε i denotes the decoding error probability of UE i. Therefore, the total data rate at UE i is:

[0020]

[0021] The sum of the data rates of all UEs is:

[0022]

[0023] Based on the above formula, the resource allocation problem can be constructed as:

[0024]

[0025] wherein, denotes a positive integer, N0 denotes the total data amount of the URLLC service sent by the AP to the UE, the (a) and (b) formulas constrain each UE to be associated with only one slice; the (c) formula constrains each slice to be non-empty; the (d) and (e) formulas constrain the number of PRBs obtained by each slice to be an integer and at least one PRB, and all PRBs are to be allocated and cannot be idle; the (f) and (g) formulas specify the value range of the power control coefficient and constrain each AP to not exceed the total power upper limit; and the (h) formula is a hard constraint on the transmission delay of the UE.

[0026] Preferably, a successive rate lower bound maximization algorithm is used to eliminate the non-convexity of the short packet rate formula, wherein the short packet data rate R i of the UEi exists a lower bound:

[0027]

[0028] wherein λ i is a rate lower bound expansion point not less than 0.28. An auxiliary variable φ i is introduced as a lower bound of Γ i , and φ i is taken as a rate lower bound expansion point. For each update of the lower bound, given a set of updated iteration variables {φ′ i} and obtained optimization variables {ω i,s}, {δ s}, {ξ i,m} of the last round of iteration, let λ i = φ′ i , and the data rate lower bound is substituted into the resource allocation problem, and the following can be obtained:

[0029]

[0030] The problem is a convex optimization problem, and the SDPT3 solver of the MATLAB CVX toolbox can be used to solve the optimal {φ i} of the current iteration, and the convex lower bound of the short packet data rate is updated.

[0031] Preferably, the SCA algorithm first introduces an auxiliary variable as a lower bound of the square root of the signal-to-interference ratio, and thus:

[0032]

[0033] Let The left side of the inequality is a non-convex function, and the right side is a convex function, so for the left side function, the first-order Taylor expansion is used for approximation, The first-order Taylor expansion of the left side of the inequality is:

[0034]

[0035] where is the first order Taylor expansion point, which is updated by iterative loop. The non-convexity of SINR is eliminated by substituting the first order Taylor expansion into the SINR lower bound constraint.

[0036] Preferably, the alternating optimization method first fixes the AP-UE associated power allocation factor {ξ i,m}, and gives the optimal {φ i} of the current iteration of the successive rate lower bound maximization algorithm and the optimal first order Taylor expansion point The original resource allocation optimization problem is transformed into:

[0037]

[0038] where and define Relaxing equation (b) as δ s > 0, The problem becomes a convex optimization problem, and the optimal continuous form of {δ s} can be solved by the SDPT3 solver of MATLAB CVX toolbox. The optimal discrete form of {δ s} can be obtained by rounding.

[0039] Then fix the number of PRBs allocated to each slice {δ s}, and define The original resource allocation optimization problem is transformed into:

[0040]

[0041] The problem is a convex problem of quadratic constraint quadratic programming, and the optimal {υ i} and {η i,m} can be solved by the SDPT3 solver of MATLAB CVX toolbox. Then update the {ξ i,m} of the current iteration of the SCA algorithm and the first order Taylor expansion point for the next iteration.

[0042] Preferably, the simulated annealing algorithm is performed after updating the PRB allocation and power allocation. After optimizing the user slice associated variable {ω i,s}, return to the next iteration of the algorithm of the successive rate lower bound maximization algorithm. The loop terminates until the network and rate change values of two iterations are lower than the threshold value.

[0043] Finally, the optimal resource allocation scheme can be obtained, and the maximum sum rate of the network can be calculated according to the optimal resource allocation scheme.

[0044] Preferably, the method is applied to smart grid control type services, and supports millisecond-level latency and nearly 100% transmission reliability in a wide temperature range of-40 to 60 DEG C.

[0045] The application further provides a communication system adopting the network slice resource allocation method, comprising a CF-mMIMO network architecture, a URLLC service support module and a resource allocation optimization module.

[0046] Compared with the prior art, the network slice resource allocation method with deterministic quality of service guarantee has the following technical effects:

[0047] The network slice resource allocation method with deterministic quality of service guarantee is used in a multi-slice user-centered CF-mMIMO network, and aims to maximize the sum rate of all users and considers the deterministic QoS demand of UEs to construct a network slice resource allocation problem, eliminate non-convexity through a successive rate lower bound maximization algorithm and an SCA algorithm, solve optimization variables through an alternating optimization method and a simulated annealing algorithm, and finally obtain an optimal resource allocation scheme. Compared with a traditional network slice resource allocation method, the application considers a new network scenario of 6G, adopts a short packet data rate formula corresponding to a URLLC service which is more in line with reality, takes the QoS demand of UEs as a constraint condition when constructing an optimization problem, and has a faster convergence speed and a higher network sum rate compared with a traditional network slice resource allocation algorithm. Therefore, the application can obtain a resource allocation scheme with higher resource utilization. BRIEF DESCRIPTION OF DRAWINGS

[0048] Figure 1 It is a multi-slice user-centered CF-mMIMO network architecture diagram.

[0049] Figure 2 It is a flowchart of a network slice resource allocation method with deterministic quality of service guarantee.

[0050] Figure 3 It is a specific implementation process of a successive rate lower bound maximization algorithm.

[0051] Figure 4 It is a specific implementation process of an SCA algorithm.

[0052] Figure 5 It is a specific implementation process of a simulated annealing algorithm. DETAILED DESCRIPTION

[0053] The specific embodiments of the present application are further described in detail below with reference to the accompanying drawings and 6G network scenarios.

[0054] The present embodiment provides services for URLLC services of users in a user-centered CF-mMIMO network architecture. A large number of APs M={1,...,m,...,M} and UEs U={1,...,i,...,U} are distributed in the network, all APs are connected to a CPU, the maximum transmission power of each AP is ρ0, the number of antennas is P, and ξ i,m represents the power allocation coefficient of AP m allocated to UE i, ξ i,m is a continuous variable between 0 and 1. UEs are allocated to different slices, and the slice set is S={1,...,s,...,S}. Each UE can only be associated with one slice, all slices are non-empty, and the association between UEs and slices is represented by 0-1 variable ω i,s . If UE i is in slice s, ω i,s =1, otherwise ω i,s =0. The bandwidth resource is divided into K PRBs with equal bandwidth, and the bandwidth of each PRB is W. Each PRB can only be used by all UEs in one slice, and the number of PRBs allocated to slice s is represented by δ s . All APs transmit data to connected UEs, and interference occurs between UEs in the same slice. UEs and APs communicate through URLLC data packets, and the maximum transmission delay that UE i can tolerate is t i,b . In order to meet the QoS requirements of UEs, the transmission delay t i of UE i realized by the network cannot exceed t i,b . In order to meet the QoS requirements of all UEs under the condition of limited spectrum resources and power resources, and on this basis to maximize the sum rate of all users, the following steps are needed to achieve:

[0055] Step 1: Initialize the CF-mMIMO network, and UEs build a set of APs serving them.

[0056] (1) The channel between each AP and UE is independent of each other, and the channel parameter h i,m is used to represent the channel fading between UE i and AP m. The channel parameter is defined as:

[0057]

[0058] where α i,m , β i,m and d i,m represent the large-scale fading parameter, path loss and distance between UE i and AP m, respectively, and β i,mis a function of d i,m i,m is a random variable subject to lognormal distribution, α i,m is defined as the product of ψ i,m and β i,m . g i,m is a small-scale fading parameter, h i,m and g i,m are Px1 vectors, each element of g i,m is a random variable subject to complex Gaussian distribution with zero mean and unit variance.

[0059] (2) In the CF-mMIMO network, the UE selects the set of APs serving it by measuring the large-scale channel gain, selects the APs satisfying the channel gain no less than a preset threshold, and selects the AP with the largest gain if there is no such AP. The construction rule of the serving set of the UE is:

[0060] C i ={m|α i,m ≥β i,m (d0)}∪{arg max m∈M α i,m}

[0061] where C i denotes the set of APs serving UE i, and d0 is a preset distance threshold of path loss. Each AP determines the connection relationship with the UE based on the serving set of the UE, and uses a 0-1 variable c i,m to represent the connection relationship between UE i and AP m. If there is a connection, c i,m = 1, otherwise c i,m = 0.

[0062] Step 2: Based on the QoS requirement of the UE, an optimization problem of resource allocation is constructed to maximize the sum rate of all users.

[0063] (1) In the CF-mMIMO network, the AP periodically sends signals to the UE, and each period is called a coherent time slot, and the period is τ c . Each coherent time slot is divided into two stages of pilot training and downlink data transmission, the pilot sequence length is τ, and the number of symbols of pilot training in a coherent time slot is also τ, and the number of symbols of downlink data transmission in a coherent time slot is τ c - τ. There are τ pairs of orthogonal pilot sequences.

[0064] ​The transmit power of each UE is p0, different pilot sequences are allocated to UEs of the same slice, and UEs of different slices can reuse the same pilot sequence. The UE transmits a pilot signal to the AP, and the AP processes the received pilot signal to predict the channel parameters between the connected UE. Assuming that the white noise at all APs obeys a Gaussian distribution with zero mean and unit variance, the independent channel gain on a single antenna between UE i and AP m can be obtained as:

[0065]

[0066] (2) For the downlink data transmission phase of the coherent time slot, all APs send data signals to the UEs connected thereto, process the symbol data and send by conjugate beamforming technology. Since different slices use different PRBs, different slices are isolated in frequency, so only the signals sent to users in the same slice will interfere with each other. Assuming that the white noise at all UEs obeys a Gaussian distribution with zero mean and unit variance, the signal-to-interference-and-noise ratio at UE i on a single PRB can be calculated as:

[0067]

[0068] Based on the formula of Γ i , the achievable short packet data rate of UE i performing URLLC service on a single PRB can be obtained as:

[0069]

[0070] Where n represents the URLLC data packet block length, Q -1 (·) represents the inverse function of the Gaussian Q function, and ε i represents the decoding error probability of UE i. Therefore, the total data rate at UE i can be obtained as:

[0071]

[0072] The sum of the data rates of all UEs is:

[0073]

[0074] (3) According to the above, the resource allocation problem can be constructed with the objective function of maximizing the system capacity as:

[0075]

[0076] Where, N0 represents the total data amount of URLLC service sent by AP to UE, (a), (b) formula restricts that each UE can only be associated with one slice; (c) formula restricts that each slice is not empty; (d), (e) formula restricts that the number of PRBs obtained by each slice is an integer and at least one PRB is obtained, and all PRBs must be allocated and cannot be idle; (f), (g) formula specifies the value range of the power control coefficient, which restricts that each AP cannot exceed the total power upper limit; (h) formula is a hard constraint on the transmission delay of UE.

[0077] Step 3: Eliminate the non-convexity of the short packet rate formula by the successive lower bound maximization algorithm.

[0078] (1) If this is the first iteration of the successive lower bound maximization algorithm, go to (2); otherwise, give the optimized {ω i,s}, {δ s}, {ξ i,m} and the signal-to-interference ratio lower bound {φ′ i} obtained in the last iteration of the successive lower bound maximization algorithm, and then go to (3).

[0079] (2) This is the first iteration of the successive lower bound maximization algorithm, initialize a set of {ω i,s}, {δ s}, {ξ i,m} and the signal-to-interference ratio lower bound {φ′ i} that satisfy the constraint conditions as the optimization variables updated in the last iteration (zeroth iteration) of the successive lower bound maximization algorithm. Initialize the maximum number of iterations I1 of the successive lower bound maximization algorithm, and the rate change value threshold and the sum rate C′ sum of the last iteration (zeroth iteration) of the successive lower bound maximization algorithm.

[0080] (3) For the short packet data rate R i of UE i, there is a lower bound:

[0081]

[0082] Where λ i is a rate lower bound expansion point not less than 0.28. Take φ i as the rate lower bound expansion point.

[0083] (4) Based on the optimized {ω i,s}, {δ s}, {ξ i,m} and the signal-to-interference ratio lower bound {φ′ i} obtained in the last iteration, let λ i = φ′i Substitute the data rate lower bound into the resource allocation problem, we can get:

[0084]

[0085] This problem is a convex optimization problem, which can be solved by the SDPT3 solver of MATLAB CVX toolbox to get the optimal {φ i} of this round of iteration, and assign {φ i} to the signal-to-interference-and-noise ratio lower bound {φ′ i} of the last iteration of the successive rate lower bound maximization algorithm. Update the convex lower bound of the short packet data rate.

[0086] Step 4: Eliminate the non-convexity of the signal-to-interference-and-noise ratio formula by the SCA algorithm.

[0087] (1) If this is the first iteration of the SCA algorithm, go to (2). Otherwise, given {ω i,s}, {ξ i,m} updated by the last iteration of the successive rate lower bound maximization algorithm and {φi} updated by this iteration of the successive rate lower bound maximization algorithm, and the first-order Taylor expansion point Then go to (3).

[0088] (2) This is the first iteration of the SCA algorithm, given {ω i,s}, {ξ i,m} updated by the last iteration of the successive rate lower bound maximization algorithm and {φi} updated by this iteration of the successive rate lower bound maximization algorithm, initialize the first-order Taylor expansion point The initialization operation is:

[0089]

[0090] Initialize the maximum number of iterations I2 of the SCA algorithm, the target function change value threshold And the objective function value V′ = 0 of the last iteration (zeroth iteration) of the SCA algorithm.

[0091] (3) Introduce an auxiliary variable As the lower bound of the square root of the signal-to-interference-and-noise ratio, so:

[0092]

[0093] Let The left side of the inequality is a non-convex function, and the right side is a convex function, so for the left function, use its first-order Taylor expansion for approximation, The first-order Taylor expansion of

[0094]

[0095] where is the first order Taylor expansion point, which is updated by iterative loop. The non-convexity of SINR is eliminated by substituting the first order Taylor expansion into the SINR lower bound constraint.

[0096] Step 5: Optimize the frequency resource and power resource respectively by the alternate optimization method.

[0097] (1) Fix the power allocation coefficient {ξ i,m} of AP-UE association, and give the optimal {φ i} of the current iteration of the successive rate lower bound maximization algorithm and the optimal first order Taylor expansion point The original resource allocation optimization problem is transformed into:

[0098]

[0099] where and define Relax (b) as δ s > 0, The problem becomes a convex optimization problem, and the optimal continuous form of {δ s} can be solved by the SDPT3 solver of MATLAB CVX toolbox. The optimal discrete form of {δ s} can be obtained by rounding.

[0100] (2) Fix the PRB number allocated to each slice {δ s}, and define The original resource allocation optimization problem is transformed into:

[0101]

[0102] The problem is a convex problem of quadratic constraint quadratic programming, and the optimal {υ i} and {η i,m} can be solved by the SDPT3 solver of MATLAB CVX toolbox, and then the {ξ i,m} of the current iteration of the SCA algorithm and the first order Taylor expansion point for the next iteration are updated. is assigned to the first order Taylor expansion point

[0103] (3) Calculate the optimal objective function value V of the quadratic constraint quadratic programming problem of the current iteration of the SCA algorithm, and judge Is it true? If true, jump back to step 4; if false, optimize {δ} in this iteration of the SCA algorithm. s},{ξ i,m} as the successive rate lower bound maximization algorithm, in this round of iterative optimization, {δ s},{ξ i,m Then proceed to step 6.

[0104] Step 6: Optimize the user slice association variables using the simulated annealing algorithm.

[0105] (1) Given the successive rate lower bound maximization algorithm, the {δ} is optimized in this round of iteration. s},{ξ i,m Initialize the initial temperature T0, and terminate the temperature T. min cooling rate α t And the maximum number of iterations I per temperature t Randomly initialize {ω′ i,s Ensure that there is at least one UE in each slice. Calculate the sum rate C at this point. sum (ω′ i,s Initialization optimal Current temperature T = T0. Cycle I at one temperature. t Second-rate.

[0106] (2) By randomly generating a neighborhood solution The neighborhood solution generation operation is as follows: with a 50% probability, randomly transfer a UE from one slice to another; with a 50% probability, randomly swap one UE in each of the two slices. When generating a new solution, it is necessary to ensure that each slice is non-empty. The sum rate corresponding to the neighborhood solution is calculated.

[0107] (3) If Then Assigned to {ω′ i,s}、C sum (ω′ i,s If the condition is not met, proceed to step (5); otherwise, proceed to step (4).

[0108] (4) A random number between 0 and 1 is generated. If the number is less than 1, then... Then Assigned to {ω′ i,s}、C sum (ω′ i,s ).

[0109] (5) If the current temperature iteration number has not reached I t If yes, then return to (2); otherwise, check C. sum (ω′i,s ) greater than whether it is true, if true, then {ω′ i,s}, C sum (ω′ i,s ) is assigned to

[0110] (6) update the current temperature T = T x a t . Determine whether T < T min is true, if true, go to (7); otherwise, return to (2).

[0111] (7) assign to {ω i,s} optimized by the current iteration of the successive lower bound maximization algorithm.

[0112] Step 7: Calculate the maximum sum rate according to the optimal allocation result.

[0113] (1) Given {δ s}, {ξ i,m} and {ω i,s} optimized by the current iteration of the successive lower bound maximization algorithm. Calculate the sum rate C sum . Determine whether is true, if true, then C sum is assigned to C′ sum , and jump back to step 3; otherwise, go to (2).

[0114] (2) Take {δ s}, {ξ i,m} and {ω i,s} optimized by the current iteration of the successive lower bound maximization algorithm as the optimal allocation result, and calculate the maximum sum rate

[0115] This embodiment is aimed at the 6G network scenario, based on the CF-mMIMO network with user as the center, aiming to maximize the sum rate of all users, combined with the deterministic QoS demand of UE, by constructing the network slice resource allocation problem, an optimization scheme is proposed. First, the successive lower bound maximization algorithm and the SCA algorithm are used to eliminate the non-convexity in the problem, and the optimization variables are solved by the alternating optimization method and the simulated annealing algorithm, and finally the optimal resource allocation scheme is realized. Compared with the traditional network slice resource allocation method, the deterministic QoS demand of UE is taken as the constraint condition, and the short packet data rate formula corresponding to the URLLC business in the actual scene is adopted; through algorithm design, the resource allocation scheme has faster convergence speed, significantly improves the network sum rate, effectively improves the resource utilization, and provides an efficient and reliable solution for network slice resource allocation in 6G network scenario.

[0116] Another embodiment is a communication system employing the network slice resource allocation method, comprising a CF-mMIMO network architecture, a URLLC service support module and a resource allocation optimization module.

[0117] The above described embodiments of the present specification have been described. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than the order in which they are recited and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve the desired results. In certain implementations, multitasking and parallel processing can be advantageous.

Claims

1. A method for allocating network slice resources with deterministic quality of service assurance, characterized in that, Includes the following steps: Construct a multi-slice, user-centric CF-mMIMO network architecture and define the connection relationship between access points (APs) and user equipment (UEs) in the CF-mMIMO network architecture; Based on the connection relationship, the QoS requirements of each UE are determined, and a resource allocation optimization problem is constructed with the goal of maximizing the sum rate of all users. The non-convexity of the short packet rate formula in the resource allocation optimization problem model is eliminated by maximizing the successive rate lower bound algorithm. The continuous convex approximation (SCA) algorithm is used to eliminate the non-convexity of the signal-to-interference-plus-noise ratio (SINR) formula in the resource allocation optimization problem model; The optimal resource allocation result is obtained by using an alternating optimization method to optimize the allocation of bandwidth and power resources respectively. The optimal user slice association result is obtained by optimizing the user slice association variables using the simulated annealing algorithm. The maximum sum rate is calculated based on the optimal resource allocation result and the optimal user slice association result.

2. The method according to claim 1, characterized in that, In the CF-mMIMO network: Bandwidth resources are divided into physical resource blocks (PRBs) of equal bandwidth, and each PRB is used by only one UE within a slice; The UE selects the set of serving APs by measuring the large-scale channel gain. If the channel gain does not reach the threshold, the UE selects the AP with the largest gain. The QoS requirement is defined as the UE's transmission latency not exceeding a preset tolerance limit.

3. The method according to claim 1, characterized in that, The successive rate lower bound maximization algorithm includes: Iteratively update the convex lower bound of the short packet rate, wherein the convex lower bound is a convex function with respect to the lower bound of the signal-to-interference-plus-noise ratio; Each update to the convex lower bound is obtained by solving a convex optimization problem with it as the optimization variable; The convex function of the lower bound of the signal-to-interference-plus-noise ratio is solved using the SDPT3 solver in the MATLAB CVX toolbox.

4. The method according to claim 1, characterized in that, The SCA algorithm introduces an auxiliary variable as a lower bound of the signal-to-interference-plus-noise ratio (SINR) after taking the square root, and uses a first-order Taylor expansion to eliminate non-convexity.

5. The method according to claim 1, characterized in that, The alternating optimization method includes: Fixed power allocation coefficients associated with user slices, optimized the number of PRBs allocated to each slice; The number of PRBs allocated to each slice is fixed, and the power allocation coefficient is optimized. The two optimization problems were solved using the SDPT3 solver in the MATLAB CVX toolbox.

6. The method according to claim 1, characterized in that, The simulated annealing algorithm optimizes user slice association variables through a neighborhood solution generation strategy, including randomly transferring or exchanging UEs within a slice, to ensure that each slice is not empty.

7. The method according to claim 1, characterized in that, The method is applied to smart grid control services, supporting millisecond-level latency and near 100% transmission reliability over a wide temperature range of -40 to 60°C.

8. A communication system, characterized in that, The network slicing resource allocation method according to any one of claims 1 to 7 includes a CF-mMIMO network architecture, a URLLC service support module, and a resource allocation optimization module.

Images