A multi-service coexistence method and device for multi-cell scenarios of power Internet of Things

Abstract

The present invention discloses a multi-service coexistence method and device for a multi-cell scenario of an electric power Internet of Things, including: constructing a multi-service coexistence implementation mode in a multi-cell scenario; modeling decision variables for user association and resource allocation of multi-service coexistence; defining user downlink rates and constraints based on decision variable modeling, and constructing a multi-objective optimization problem about the downlink rate and of all eMBB users in each cell; solving the multi-objective optimization problem under constraints to obtain a Pareto solution set that maximizes the downlink rate and of all eMBB users in each cell, and obtaining the optimal decision variables for user association and resource allocation as a load balancing solution based on the Pareto solution set to achieve multi-service coexistence deployment. The present invention can ensure the load balancing of each cell while ensuring the low latency and high reliability requirements of the uRLLC service, while maximizing the rate and of all cells, and improving the throughput of the eMBB service.

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and specifically relates to a multi-service coexistence method and device for a multi-cell scenario of an electric power Internet of Things. Background Art

[0002] Since the implementation of 5G (fifth generation mobile communication technology), the speed and reliability of wireless communication data transmission have been further improved, and the latency has been further reduced, driving the explosive development of smart medical care, smart driving and other fields. In order to support the performance indicators that 5G needs to achieve, three major application scenarios of 5G are defined, including enhanced mobile broadband (eMBB), ultra-reliable low latency communication (uRLLC) and massive machine type communication (mMTC). Among them, enhanced mobile broadband can greatly increase the peak data rate, accommodate more traffic, and serve more users; ultra-reliable low latency communication will control the connection latency to 1 millisecond level, and support high-reliability connections under high-speed movement; massive machine type communication can meet the network access needs of massive devices.

[0003] From the characteristics of the above services, it can be seen that under the control of Shannon's theorem, eMBB services require the support of high transmission bandwidth, which is reflected in the need for the communication system to allocate more resource blocks to the service. The data volume of uRLLC services is usually small, but its high reliability and low latency indicators require the communication system to support burst transmission, that is, to transmit immediately after the uRLLC service arrives, rather than waiting for other services to be transmitted before transmitting. In the case of limited communication resource blocks, how to allocate the resource blocks occupied by eMBB services and uRLLC services, while ensuring the low latency requirements of uRLLC services, and maximizing the throughput of the communication system is a major concern of the current academic and industrial circles.

[0004] In order to realize the coexistence of eMBB and uRLLC services, a simple solution is resource reservation, which ensures low latency and reliability by allocating part of the frequency band to the uRLLC service. However, this solution will cause a waste of frequency band resources and reduce the throughput of eMBB services when there is no uRLLC service. In the paper entitled "Delay-aware Dynamic Resource Allocation Scheme for 5G Heterogeneous Networks: Network Slicing-Multi-User Scenario", a repeated multiplexing scheme is proposed. This scheme multiplexes eMBB and uRLLC services in the same resource block to improve spectrum utilization and service throughput. However, there is currently no mature interference elimination algorithm to reduce the impact between different services, so this scheme has the disadvantage of reducing the reliability of uRLLC services.

[0005] The current standard solution in the industry is the puncturing technology based on short transmission time intervals proposed by the 3rd Generation Partnership Project (3GPP). Under this technical provision, when uRLLC services arrive, if there are no additional resource blocks, the resource blocks already allocated to the eMBB services will be preempted for transmission to meet the requirements of low latency and high reliability of uRLLC services. However, the way to preempt resource blocks, that is, the resource scheduling algorithm, will significantly affect the performance indicators such as the throughput of the communication system. Therefore, it is crucial to design a reasonable resource scheduling algorithm.

[0006] In related research, the Chinese patent application with publication number CN118338450A "eMBB and URLLC coexistence data flow scheduling method in 5G network based on reinforcement learning" proposes a method for scheduling eMBB and URLLC coexistence data flows in 5G network based on reinforcement learning, which guarantees the deadline requirements of URLLC data flows; guarantees the minimum data rate requirements of eMBB data flows; and maximizes the minimum fairness index in eMBB data flows. This method uses the verification ability of reinforcement learning in the offline stage and the continuous learning ability in the online stage to implement the data flow scheduling strategy. The Chinese patent application with publication number CN118413896A "Uplink eMBB and uRLLC Coexistence Scheduling Method and System Based on Random Forest" uses a hierarchical scheduling algorithm and random forest to predict the decoding success rate of eMBB users after puncturing, generates the throughput loss of eMBB services based on the decoding success rate, calculates the puncturing priority of each eMBB user based on the throughput loss of the eMBB service and the number of puncturing times of each user, and formulates the puncturing strategy of uRLLC traffic based on the puncturing priority to achieve uplink eMBB and uRLLC coexistence scheduling.

[0007] The above-mentioned related technologies focus on the eMBB and uRLLC service allocation algorithms in single-cell scenarios. In actual communication scenarios, users are often within the service range of multiple base stations, so they can choose to send data packets to multiple base stations. For example, when the base station service load is uneven, by sending uRLLC data packets to idle base stations, the impact on the eMBB services of all cells can be minimized while ensuring the service demand. Therefore, it is of great significance to study the scheduling algorithm for the coexistence scenario of eMBB and uRLLC services in multi-cell situations. However, the scheduling algorithm for multiple cells usually involves a non-convex mixed integer nonlinear programming problem, which has the characteristics of high computational complexity and easy convergence to a local optimal solution, and the optimization problem is difficult to solve. Summary of the invention

[0008] In view of the above, the purpose of the present invention is to provide a multi-service coexistence method and device for a multi-cell scenario of the power Internet of Things. Aiming at the scheduling problem of the coexistence of eMBB and uRLLC services in the power heterogeneous network, a coexistence service resource scheduling algorithm in a multi-cell scenario is designed using the NSGA-Ⅲ algorithm (non-dominated sorting genetic algorithm III). In this scenario, users can select a base station that provides service services according to the current load of each base station. Compared with the existing technical solutions, the present invention can ensure the load balancing of each cell while ensuring the low latency and high reliability requirements of the uRLLC service, while maximizing the rate of all cells and improving the throughput of the eMBB service.

[0009] In order to achieve the above-mentioned invention object, the technical solution provided by the present invention is as follows:

[0010] In a first aspect, an embodiment of the present invention provides a multi-service coexistence method for a multi-cell scenario of an electric power Internet of Things, comprising the following steps:

[0011] Construct a multi-service coexistence implementation mode including eMBB services and uRLLC services in multi-cell scenarios;

[0012] In the implementation mode, decision variables for user association and resource allocation for coexisting multiple services are modeled. The decision variables for user association include the base station selection strategy for eMBB users and the base station selection strategy for uRLLC users. The decision variables for resource allocation include the resource block allocation method, the puncturing density of uRLLC users in different resource blocks, and the base station's transmit power allocation for eMBB users.

[0013] Define the downlink rate obtained by eMBB users in any cell, the achievable downlink rate obtained by uRLLC users in any cell, the high reliability and low latency constraints of uRLLC services constructed based on the achievable downlink rate obtained by uRLLC users in any cell, the base station power constraints, and the base station resource block constraints according to the decision variable modeling, calculate the sum of the downlink rates of all eMBB users in any cell, and construct a multi-objective optimization problem on the sum of the downlink rates of all eMBB users in each cell;

[0014] Based on the multi-objective optimization problem, the downlink rate of all eMBB users in each cell is solved under constraints to obtain the maximum Pareto solution set. Based on the Pareto solution set, the optimal decision variables for user association and resource allocation are obtained as the load balancing solution, and multi-service coexistence deployment is achieved based on the load balancing solution.

[0015] Preferably, the construction of a multi-service coexistence implementation mode including eMBB services and uRLLC services in a multi-cell scenario includes:

[0016] Based on the perforated scheduling strategy, in the scenario where eMBB services and uRLLC services coexist in multiple cells, when a uRLLC service burst occurs, several resource blocks across frequency bands are selected from the time-frequency resource blocks of the eMBB service, and the symbols in these resource blocks are replaced with the symbols of the uRLLC data packets to ensure the uRLLC service. The replaced resource blocks will be retransmitted to ensure the completion of the eMBB service.

[0017] Preferably, in the implementation mode, decision variable modeling is performed for user association and resource allocation for coexistence of multiple services, the decision variables for user association include the base station selection strategy for eMBB users and the base station selection strategy for uRLLC users, and the decision variables for resource allocation include the resource block allocation method, the puncturing density of uRLLC users in different resource blocks, and the base station's transmit power allocation to eMBB users, including:

[0018] The coverage of a base station is regarded as a cell. In the implementation mode, there are C cells in a region, and there are overlapping areas between the cells. There are M eMBB users and N uRLLC users randomly distributed in the overlapping coverage of these cells. These two types of users can select a base station.

[0019] The decision variables for user association include the base station selection strategy a for eMBB users. m,c and base station selection strategy a for uRLLC users n,c , each user can only choose one base station to connect to, satisfying

[0020] The decision variables for resource allocation include the resource block allocation method x m,b , uRLLC user puncturing density in different resource blocks s b (t) and the transmit power allocation p of the eMBB user m,b,c (t); where resource block allocation mode x m,b is the associated variable representing the eMBB user m occupying resource block b, satisfying x m,b ∈{0,1}, uRLLC user puncturing density s in different resource blocks b (t) is the number of mini-slots occupied by uRLLC services in resource block b, satisfying s b (t)∈{0,1,2,...,K-1,K}, K is the maximum number of schedulable time slots; the transmit power allocation p of the eMBB user m,b,c (t) is the power allocated to eMBB user m by cell c on resource block b.

[0021] Preferably, the calculation formula of the downlink rate obtained by the eMBB user in any cell is expressed as:

[0022]

[0023] Among them, h m,b,c (t) represents the channel power gain of cell c on resource block b for eMBB user m within time slot t, W represents the bandwidth of the resource block, σ 2 represents the noise power;

[0024] The calculation formula for the achievable downlink rate that a uRLLC user can obtain in any cell is expressed as:

[0025]

[0026] Among them, Q -1 represents the inverse Gaussian distribution, represents the transmission error probability, D n,b (t) represents the channel dispersion characteristic, which determines the random variation degree of the channel belonging to resource block b of the uRLLC user within the range of time slot t, p n,b,c (t) represents the transmission power of uRLLC user, h n,b,c (t) represents the channel gain of uRLLC user, σ 2 represents the noise power, l n,b (t) represents the length of the storage unit in the symbol.

[0027] Preferably, the high reliability and low latency constraints of the uRLLC service constructed based on the achievable downlink rate obtained by the uRLLC user in any cell are expressed as:

[0028]

[0029] Where Pr{} represents the probability, ξ represents the length of the data packet sent by the uRLLC service, L(t) represents the total number of uRLLC data packets arriving in a single time slot t, and θ max represents a given maximum probability threshold;

[0030] The base station power constraint is expressed as:

[0031]

[0032] p m,b,c (t)≥0

[0033] Among them, P max Indicates the maximum transmit power of the base station;

[0034] The base station resource block constraints are expressed as:

[0035]

[0036] Among them, Bc It represents the number of resource blocks allocated to cell c, and B represents the total number of available resource blocks in the scenario of multi-cell power heterogeneous network.

[0037] Preferably, the calculating the sum of downlink rates of all eMBB users in any cell and constructing a multi-objective optimization problem about the sum of downlink rates of all eMBB users in each cell includes:

[0038] The downlink rate and f of all eMBB users in cell c c It is expressed as:

[0039]

[0040] Then the downlink rate sum of all eMBB users in each cell is expressed as {f1,f2,…,f C}, the multi-objective optimization problem is expressed as:

[0041]

[0042] Preferably, the NSGA-Ⅲ algorithm is used to solve the multi-objective optimization problem to obtain the Pareto solution set P, {a m,c ,a n,c ,p m,b,c (t),s b (t),x m,b}∈P.

[0043] Preferably, obtaining the optimal decision variables regarding user association and resource allocation as a load balancing solution based on the Pareto solution set includes:

[0044] Based on the NSGA-Ⅲ algorithm, after solving the multi-objective optimization problem and obtaining the Pareto solution set, the following formula is further solved:

[0045]

[0046] The optimal decision variable {a m,c ,a n,c ,p m,b,c (t),s b (t),x m,b} as a load balancing solution.

[0047] In a second aspect, in order to achieve the above-mentioned invention object, an embodiment of the present invention further provides a multi-service coexistence device for a multi-cell scenario of an electric power Internet of Things, comprising: a service scenario construction module, a decision variable modeling module, an optimization problem construction module and an optimization problem solving module;

[0048] The service scenario construction module is used to construct a multi-service coexistence implementation mode including eMBB services and uRLLC services in a multi-cell scenario;

[0049] The decision variable modeling module is used to perform decision variable modeling on user association and resource allocation for coexistence of multiple services in an implementation mode, wherein the decision variables for user association include a base station selection strategy for eMBB users and a base station selection strategy for uRLLC users, and the decision variables for resource allocation include a resource block allocation method, a puncturing density of uRLLC users in different resource blocks, and a transmit power allocation for eMBB users;

[0050] The optimization problem construction module is used to define the downlink rate obtained by the eMBB user in any cell, the achievable downlink rate obtained by the uRLLC user in any cell, the high reliability and low latency constraints of the uRLLC service constructed based on the achievable downlink rate obtained by the uRLLC user in any cell, the base station power constraints, and the base station resource block constraints according to the decision variable modeling, calculate the downlink rate sum of all eMBB users in any cell, and construct a multi-objective optimization problem about the downlink rate sum of all eMBB users in each cell;

[0051] The optimization problem solving module is used to solve the Pareto solution set that maximizes the downlink rate of all eMBB users in each cell under constraints, and obtain the optimal decision variables for user association and resource allocation as a load balancing solution based on the Pareto solution set, and realize multi-service coexistence deployment based on the load balancing solution.

[0052] On the third aspect, in order to achieve the above-mentioned purpose of the invention, an embodiment of the present invention also provides an electronic device, including a memory and one or more processors, wherein the memory is used to store a computer program, and the processor is used to implement the above-mentioned multi-service coexistence method for the multi-cell scenario of the power Internet of Things when executing the computer program.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] (1) In solving the problem of coexistence of multiple services, the present invention comprehensively considers resource block allocation, puncturing density allocation, power, uRLLC service reliability and low latency constraints. On this basis, a multi-cell resource reuse model is introduced. When a user is within the service range of multiple base stations, the base station can be selected for association based on the channel status and the spectrum resources that can be allocated to each base station. While ensuring the reliability of uRLLC service transmission, the system capacity is maximized and the load balance between the cells is guaranteed.

[0055] (2) The present invention uses the NSGA-III algorithm to achieve an approximate solution to the optimal solution of the non-convex mixed integer nonlinear programming problem of user association and resource allocation for coexisting eMBB and uRLLC services in a multi-cell power heterogeneous network. The algorithm converges quickly, and the approximate solution performs better than the existing algorithms in simulation conditions.

[0056] (3) The present invention enables the power heterogeneous network to further improve the downlink rate of the eMBB service while ensuring the uRLLC service, which helps to meet the large number of communication needs in the future power heterogeneous network, further improve the communication efficiency, and enhance the economic benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0058] Figure 1 It is a flowchart of a multi-service coexistence method for a multi-cell scenario of an electric power Internet of Things provided by an embodiment of the present invention;

[0059] Figure 2 It is a schematic diagram of using a perforation scheduling strategy to achieve coexistence of eMBB and uRLLC services provided by an embodiment of the present invention;

[0060] Figure 3 It is a user distribution diagram in which eMBB users and uRLLC users can select base stations for communication according to strategies in a multi-cell power heterogeneous network provided by an embodiment of the present invention;

[0061] Figure 4 It is a flow chart of solving the multi-service coexistence problem of a multi-cell power heterogeneous network based on the NSGA-III algorithm provided by an embodiment of the present invention;

[0062] Figure 5 It is a structural schematic diagram of a multi-service coexistence device for a multi-cell scenario of an electric power Internet of Things provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0063] To make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific implementation methods described herein are only used to explain the present invention and do not limit the scope of protection of the present invention.

[0064] The inventive concept of the present invention is: in response to the scheduling problem of coexistence of eMBB and uRLLC services in a multi-cell scenario in a power heterogeneous network in the prior art, the embodiment of the present invention provides a multi-service coexistence method and device for a multi-cell scenario of the power Internet of Things, by systematically modeling the user association and resource allocation strategy of the coexistence of multiple services, and constructing a multi-objective optimization problem for solving, and using the NSGA-Ⅲ algorithm to further obtain a load balancing solution, so that users can select a base station that provides service services according to the current load of each base station, while ensuring the low latency and high reliability requirements of the uRLLC service, ensuring the load balancing of each cell, and maximizing the rate and of all cells, thereby improving the throughput of the eMBB service.

[0065] Figure 1 1 is a flow chart of a multi-service coexistence method for a multi-cell scenario of the power Internet of Things provided by an embodiment of the present invention. Figure 1 As shown, the embodiment provides a multi-service coexistence method for a multi-cell scenario of an electric power Internet of Things, including the following steps:

[0066] S1, build a multi-service coexistence implementation mode including eMBB services and uRLLC services in a multi-cell scenario.

[0067] As one of the main services of 5G, eMBB and uRLLC have a wide range of applications, so in many scenarios, the two services coexist. Among them, eMBB has the characteristics of large data transmission volume, but the requirement for latency is not high. uRLLC services usually transmit small amounts of data, but are sensitive to latency and reliability.

[0068] In a multi-cell heterogeneous network, when two services occur at the same time, the user can select a suitable base station from the base stations providing services and send service data packets to the base station. The load of eMBB services is usually large, so they will occupy most of the shared resource blocks. In this case, when uRLLC services occur, in order to ensure low latency and high reliability of uRLLC services, the base station will use a perforated scheduling strategy.

[0069] Specifically, because the eMBB downlink rate requirement is large, the base station allows the eMBB service to span multiple time resources to improve spectrum efficiency. Because of the short latency requirement for the uRLLC service, the base station allows the uRLLC service to span multiple frequency domain resources to reduce transmission latency. Therefore, from a time domain perspective, it can be considered that eMBB users are scheduled in longer time slots, while uRLLC users are scheduled in short mini-time slots. One eMBB time slot can contain several uRLLC time slots. The perforation scheduling strategy is that when a uRLLC service burst occurs, several resource blocks that span the frequency band are selected from the eMBB time-frequency resource blocks, so that the symbols in these resource blocks are replaced by the symbols of the uRLLC data packets to ensure the uRLLC service. The replaced resource blocks will be retransmitted to ensure the completion of the eMBB service. The schematic diagram of the perforation scheduling strategy is shown below. Figure 2 shown.

[0070] The base station uses puncturing to puncture the ongoing eMBB transmission to send uRLLC data packets, and the punctured eMBB data is retransmitted after a certain period of time. Under such a scheduling strategy, the transmission performance of eMBB users is reduced in exchange for the guarantee of uRLLC user performance.

[0071] S2, in the implementation mode, the decision variables of user association and resource allocation for coexistence of multiple services are modeled. The decision variables of user association include the base station selection strategy of eMBB users and the base station selection strategy of uRLLC users. The decision variables of resource allocation include the resource block allocation method, the puncturing density of uRLLC users in different resource blocks, and the base station's transmit power allocation to eMBB users.

[0072] The coverage of a base station is regarded as a cell. In the implementation mode, there are C cells in a region, and there are overlapping areas between the cells. There are M eMBB users and N uRLLC users randomly distributed in the overlapping coverage of these cells. These two types of users can select a base station. The user distribution diagram is shown in the figure below. Figure 3 .

[0073] The decision variables for user association include the base station selection strategy a for eMBB users. m,c and base station selection strategy a for uRLLC users n,c , each user can only choose one base station to connect to, satisfying

[0074] All users within the overlapping coverage area of ​​the cell share B resource blocks. In order to ensure that there is no interference between cells, different base stations cannot reuse the same resource block. In the embodiment of the present invention, the division of resource blocks is mainly based on the frequency domain, that is, different base stations use resource blocks on different subcarriers to ensure that there are resource blocks available for uRLLC services at any time t. The bandwidth of each resource block b is W, and the number of resource blocks allocated to each base station in each time slot t is B. c eMBB user m can select several resource blocks from these resource blocks to occupy, and the occupation relationship between user m and resource block b is recorded as the associated variable x m,b .

[0075] Considering the differences between eMBB and uRLLC services, the scheduling granularity of uRLLC is smaller than that of eMBB, that is, the time slot of each resource block usually contains several uRLLC scheduling time slots, and the maximum number of schedulable time slots is set to K. When a uRLLC service occurs, the base station can select s from the K optional time slots in resource block b. b (t) time slots are used for uRLLC service transmission.

[0076] The transmission power of the base station is limited, so the sum of the powers allocated to different users at the same time cannot exceed the maximum transmission power P max .

[0077] After allocating the base station selection strategy, resource block partitioning strategy, uRLLC time slot selection strategy and transmit power allocation strategy, the eMBB user downlink rate r in time slot t can be obtained. m (t), uRLLC user downlink rate r n (t). In the next section, the calculation method of these two rates will be given, and according to the design requirements of user association and resource allocation methods for coexisting multiple services, an optimization problem will be established to make these two rates reach appropriate values.

[0078] S3, based on decision variable modeling, defines the downlink rate obtained by eMBB users in any cell, the achievable downlink rate obtained by uRLLC users in any cell, the high reliability and low latency constraints of uRLLC services constructed based on the achievable downlink rate obtained by uRLLC users in any cell, the base station power constraints, and the base station resource block constraints, calculates the sum of the downlink rates of all eMBB users in any cell, and constructs a multi-objective optimization problem regarding the sum of the downlink rates of all eMBB users in each cell.

[0079] For eMBB users, their performance is mainly expressed by the rate, which is the rate expression obtained by Shannon's formula analysis. Within the time slot t range, the calculation formula for the downlink rate obtained by an eMBB user in any cell is expressed as:

[0080]

[0081] Among them, x m,b represents the association variable between eMBB user m and resource block b, satisfying x m,b ∈{0,1}, s b (t) represents the number of mini-slots occupied by uRLLC services in resource block b, s b (t)∈{0,1,2,...,K-1,K},p m,b,c (t) represents the power allocated by cell c to cell m on resource block b, h m,b,c (t) is the corresponding channel power gain, σ 2 is the noise power.

[0082] The performance of uRLLC users can no longer be described by Shannon’s formula, because the coding blocks of uRLLC services are usually limited. Considering the requirements of short data packets and ultra-low latency, the finite length channel capacity is generally used to measure the data rate of uRLLC services. Let r n (t) represents the achievable downlink rate of uRLLC user n in time slot t, expressed as:

[0083]

[0084] Among them, Q -1 represents the inverse Gaussian distribution, represents the transmission error probability, D n,b (t) represents the channel dispersion characteristics, and the calculation formula will be given below, p n,b,c (t) represents the transmission power of uRLLC user, h n,b,c (t) represents the channel gain of uRLLC user, σ 2 represents the noise power, l n,b (t) represents the length of the storage unit (corydonbyte, CB) in the symbol.

[0085] D n,b (t) determines the random variation degree of the channel of user n belonging to resource block b at time point t, satisfying the following formula:

[0086]

[0087] In the embodiment, the performance of the system is considered in three aspects. The first is the performance of uRLLC. Since it has the highest priority, its success rate needs to be guaranteed first. The second is the performance of eMBB users. Generally, the rate needs to be as high as possible. Finally, the load balancing of different cells. In order to avoid a cell being overloaded or too idle, the load of all cells needs to be maintained at the same level. Its load is mainly reflected in the rate and of the eMBB users it serves.

[0088] For the performance of uRLLC users, it is necessary to ensure that the interruption probability of all uRLLC services in any time slot t is less than a given threshold θ max , usually this value is much less than 1.

[0089] Assuming that the total number of uRLLC data packets arriving in a single time slot t is L(t), the high reliability and low latency constraints of the uRLLC service built based on the achievable downlink rate obtained by the uRLLC user in any cell are expressed as:

[0090]

[0091] Wherein, Pr{} represents the probability, and ξ represents the length of the data packet sent by the uRLLC service.

[0092] The base station power constraints must also be met:

[0093]

[0094] p m,b,c (t)≥0

[0095] Among them, P max Indicates the maximum transmit power of the base station.

[0096] Since different base stations cannot reuse resource blocks, the total resource blocks are limited, and the resource blocks allocated to each base station are also limited, so the base station resource block constraints are also included:

[0097]

[0098] Among them, B c It represents the number of resource blocks allocated to cell c, and B represents the total number of available resource blocks in the scenario of multi-cell power heterogeneous network.

[0099] In order to achieve load balancing among cells in subsequent calculations, it is necessary to calculate the downlink rate and f of all eMBB users in cell c. c , expressed as:

[0100]

[0101] Then the downlink rate sum of all eMBB users in each cell is expressed as {f1,f2,…,f C In the embodiment, a multi-objective optimization problem is set to solve the Pareto solution set P that maximizes the sum of the eMBB user rates of each cell, {a m,c ,a n,c ,p m,b,c (t),s b (t),x m,b}∈P, to maximize the downlink rate of eMBB services. In the next section, in order to take into account the load balancing between cells, it is necessary to find the Pareto solution set P obtained. This solution is taken as the power balancing solution.

[0102] The complete multi-objective optimization problem is expressed as:

[0103]

[0104] The following constraints are met:

[0105]

[0106] S4, based on the multi-objective optimization problem, solve the downlink rate of all eMBB users in each cell under the constraints to obtain the maximum Pareto solution set, and on the basis of the Pareto solution set, obtain the optimal decision variables for user association and resource allocation as the load balancing solution, and realize the coexistence deployment of multiple services based on the load balancing solution.

[0107] The optimization problem in this embodiment is a non-convex mixed integer nonlinear programming problem, and also a multi-objective optimization problem. Solving the optimal solution of this type of problem is an NP-hard problem, and it is difficult for conventional search algorithms to find the optimal solution. For this reason, the NSGA-III algorithm is used to obtain an approximate solution to the problem.

[0108] NSGA-Ⅲ (Non-dominated Sorting Genetic Algorithm III) is a search algorithm for solving multi-objective optimization problems. It is an improved version of the NSGA-II algorithm, especially suitable for dealing with problems with multiple (usually three or more) objectives. The algorithm maintains the diversity of the population by introducing the concept of reference points, and can more efficiently search for solutions in high-dimensional objective spaces.

[0109] In order to use the NSGA-III algorithm to solve the optimization problem in this embodiment, the decision variables must first be encoded. The decision variables of this problem include {a m,c ,a n,c ,p m,b,c (t),s b (t),x m,b}, except p m,b,cThe decision variables outside (t) belong to discrete finite sets. According to the constraints, the encoding method is as follows:

[0110] According to the constraints and a m,c ∈{0,1}, a m,c The encoding is a one-dimensional vector of length M, and each component is an integer between [1, C].

[0111] According to the constraints and a n,c ∈{0,1}, a n,c The encoding is a one-dimensional vector of length N, and each component is an integer between [1, C].

[0112] According to the constraints and x m,b ∈{0,1}, and x m,b The encoding is a one-dimensional vector of length B, where each component is an integer between [1, M].

[0113] According to the constraints b (t)∈{0,1,2,...,K-1,K}, and s b (t) is encoded as a one-dimensional vector of length B, where each component is an integer between [0, K].

[0114] According to a m,c ∈{0,1} and p m,b,c (t)≥0, the constraint Strengthen to On this basis, p m,b,c (t) is encoded as C groups of one-dimensional vectors of length M×B, all components are between [0,1], and the sum of the components of each group of vectors is 1.

[0115] The above decision variables are encoded and reorganized as the decision variables of NSGA-III for optimization. The process is as follows: Figure 4 As shown, the specific steps are as follows:

[0116] (1) Parameter initialization: Randomly generate the initial population corresponding to the decision variables.

[0117] (2) Non-dominated sorting: The population is sorted according to the dominance relationship between individuals to form different non-dominated layers.

[0118] (3) Reference point generation: Generate a set of predefined reference points in the target space.

[0119] (4) Crossover and mutation: Generate new individuals corresponding to the decision variables through genetic operators.

[0120] (5) Population update: Combine the parent and child generations and check whether they meet the uRLLC low latency constraint. If not, eliminate them to form a new population.

[0121] (6) Selection operation: Based on the non-dominated sorting and the reference point, select individuals from the new population to form the next generation population. Return to step (2).

[0122] (7) Termination: When the number of iterations reaches the set value, the iteration stops and the Pareto optimal solution is output.

[0123] (8) Selection of load balancing solution: According to the formula Find the decision variables that maximize the downlink rate of eMBB users in each cell and ensure that the downlink rates provided by each base station are as close as possible, that is, achieve a load balancing solution between base stations as the result of the optimization algorithm.

[0124] After obtaining the solution of the optimization problem, the decision variables are decoded as follows:

[0125] The decision variable is a one-dimensional vector x with a length of M+N+2B+C*M*B opt , slice it.

[0126] x opt Each component of [1:M] is an integer between [1,C]. m,c The value of satisfies:

[0127]

[0128] x opt Each component of [M+1:M+N] is an integer between [1,C]. n,c The value of satisfies:

[0129]

[0130] x opt Each component of [M+N+1:M+N+B] takes an integer between [1,M]. m,b The value of satisfies:

[0131]

[0132] x opt Each component of [M+N+B+1:M+N+2B] is an integer between [0,K]. b The value of (t) satisfies:

[0133] s b (t) = x opt [M+N+B+b]

[0134] Transmit power p m,b,c (t) Obey

[0135] p m,b,c (t) = P max *x opt [M+N+2B+(c-1)*M*B+(b-1)*M+m]

[0136] Using the decoding result as the user association and resource allocation strategy for time slot t can maximize the eMBB downlink rate and meet the load balancing of each cell while meeting the low latency and high reliability requirements of the uRLLC service.

[0137] In summary, the multi-service coexistence method for the multi-cell scenario of the power Internet of Things provided by the embodiment of the present invention can maximize the downlink rate and of eMBB users in each cell while ensuring the high reliability and low latency requirements of the uRLLC service, while ensuring that the downlink rate provided by each base station is as close as possible, that is, achieving load balancing between base stations.

[0138] Based on the same inventive concept, Figure 5 As shown, an embodiment of the present invention also provides a multi-service coexistence device 500 for a multi-cell scenario of an electric power Internet of Things, including: a service scenario construction module 510, a decision variable modeling module 520, an optimization problem construction module 530 and an optimization problem solving module 540.

[0139] The service scenario construction module 510 is used to construct a multi-service coexistence implementation mode including eMBB services and uRLLC services in a multi-cell scenario.

[0140] The decision variable modeling module 520 is used to perform decision variable modeling on user association and resource allocation for coexistence of multiple services in the implementation mode. The decision variables of user association include the base station selection strategy of eMBB users and the base station selection strategy of uRLLC users. The decision variables of resource allocation include the resource block allocation method, the puncturing density of uRLLC users in different resource blocks and the base station's transmission power allocation to eMBB users.

[0141] The optimization problem construction module 530 is used to define the downlink rate obtained by eMBB users in any cell, the achievable downlink rate obtained by uRLLC users in any cell, the high reliability and low latency constraints of uRLLC services constructed based on the achievable downlink rate obtained by uRLLC users in any cell, the base station power constraints, and the base station resource block constraints according to decision variable modeling, calculate the sum of the downlink rates of all eMBB users in any cell, and construct a multi-objective optimization problem regarding the sum of the downlink rates of all eMBB users in each cell.

[0142] The optimization problem solving module 540 is used to solve the Pareto solution set that maximizes the downlink rate of all eMBB users in each cell under constraints, and obtain the optimal decision variables for user association and resource allocation as a load balancing solution based on the Pareto solution set, and realize multi-service coexistence deployment based on the load balancing solution.

[0143] Based on the same inventive concept, an embodiment of the present invention also provides an electronic device, including a memory and one or more processors, the memory is used to store a computer program, and the processor is used to implement the above-mentioned multi-service coexistence method for the multi-cell scenario of the power Internet of Things when executing the computer program.

[0144] It should be noted that the multi-service coexistence device and electronic device for multi-cell scenarios of the power Internet of Things provided in the above-mentioned embodiments belong to the same inventive concept as a multi-service coexistence method for multi-cell scenarios of the power Internet of Things. The specific implementation process is detailed in an embodiment of a multi-service coexistence method for multi-cell scenarios of the power Internet of Things, which will not be repeated here.

[0145] The specific implementation methods described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above is only the most preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A multi-service coexistence method for a multi-cell scenario of the power Internet of Things, characterized in that: The following steps are involved: Construct a multi-service coexistence implementation mode including eMBB services and uRLLC services in multi-cell scenarios; In the implementation mode, decision variables for user association and resource allocation for coexisting multiple services are modeled. The decision variables for user association include the base station selection strategy for eMBB users and the base station selection strategy for uRLLC users. The decision variables for resource allocation include the resource block allocation method, the puncturing density of uRLLC users in different resource blocks, and the base station's transmit power allocation for eMBB users. Define the downlink rate obtained by eMBB users in any cell, the achievable downlink rate obtained by uRLLC users in any cell, the high reliability and low latency constraints of uRLLC services constructed based on the achievable downlink rate obtained by uRLLC users in any cell, the base station power constraints, and the base station resource block constraints according to the decision variable modeling, calculate the sum of the downlink rates of all eMBB users in any cell, and construct a multi-objective optimization problem on the sum of the downlink rates of all eMBB users in each cell; Based on the multi-objective optimization problem, the downlink rate of all eMBB users in each cell is solved under constraints to obtain the maximum Pareto solution set. Based on the Pareto solution set, the optimal decision variables for user association and resource allocation are obtained as the load balancing solution, and multi-service coexistence deployment is achieved based on the load balancing solution.

2. The multi-service coexistence method for multi-cell scenarios of the power Internet of Things according to claim 1 is characterized in that: The construction of a multi-service coexistence implementation mode including eMBB services and uRLLC services in a multi-cell scenario includes: Based on the perforated scheduling strategy, in the scenario where eMBB services and uRLLC services coexist in multiple cells, when a uRLLC service burst occurs, several resource blocks across frequency bands are selected from the time-frequency resource blocks of the eMBB service, and the symbols in these resource blocks are replaced with the symbols of the uRLLC data packets to ensure the uRLLC service. The replaced resource blocks will be retransmitted to ensure the completion of the eMBB service.

3. The multi-service coexistence method for multi-cell scenarios of the power Internet of Things according to claim 1 is characterized in that: In the implementation mode, decision variable modeling is performed for user association and resource allocation for coexistence of multiple services. The decision variables for user association include base station selection strategies for eMBB users and base station selection strategies for uRLLC users. The decision variables for resource allocation include resource block allocation methods, puncturing densities of uRLLC users in different resource blocks, and transmission power allocation of base stations to eMBB users, including: The coverage of a base station is regarded as a cell. In the implementation mode, there are C cells in a region, and there are overlapping areas between the cells. There are M eMBB users and N uRLLC users randomly distributed in the overlapping coverage of these cells. These two types of users can select a base station. The decision variables for user association include the base station selection strategy a for eMBB users. m,c and base station selection strategy a for uRLLC users n,c , each user can only choose one base station to connect to, satisfying The decision variables for resource allocation include the resource block allocation method x m,b , uRLLC user puncturing density in different resource blocks s b (t) and the transmit power allocation p of the eMBB user m,b,c (t); where resource block allocation mode x m,b is the associated variable representing the eMBB user m occupying resource block b, satisfying x m,b ∈{0,1}, uRLLC user puncturing density s in different resource blocks b (t) is the number of mini-slots occupied by uRLLC services in resource block b, satisfying s b (t)∈{0,1,2,...,K-1,K}, K is the maximum number of schedulable time slots; the transmit power allocation p of the eMBB user m,b,c (t) is the power allocated to eMBB user m by cell c on resource block b.

4. The multi-service coexistence method for multi-cell scenarios of the power Internet of Things according to claim 3 is characterized in that: The calculation formula for the downlink rate obtained by an eMBB user in any cell is expressed as: Among them, h m,b,c (t) represents the channel power gain of cell c on resource block b for eMBB user m within time slot t, W represents the bandwidth of the resource block, σ 2 represents the noise power; The calculation formula for the achievable downlink rate that a uRLLC user can obtain in any cell is expressed as: Among them, Q -1 represents the inverse Gaussian distribution, represents the transmission error probability, D n,b (t) represents the channel dispersion characteristic, which determines the random variation degree of the channel belonging to the resource block of the uRLLC user within the time slot t, satisfying the following formula: p n,b,c (t) represents the transmission power of uRLLC user, h n,b,c (t) represents the channel gain of uRLLC user, σ 2 represents the noise power, l n,b (t) represents the length of the storage unit in the symbol.

5. The multi-service coexistence method for multi-cell scenarios of the power Internet of Things according to claim 4 is characterized in that: The high reliability and low latency constraints of the uRLLC service built based on the achievable downlink rate that the uRLLC user can obtain in any cell are expressed as: Where Pr{} represents the probability, ξ represents the length of the data packet sent by the uRLLC service, L(t) represents the total number of uRLLC data packets arriving in a single time slot t, and θ max represents a given maximum probability threshold; The base station power constraint is expressed as: p m,b,c (t)≥0 Among them, P max Indicates the maximum transmit power of the base station; The base station resource block constraints are expressed as: Among them, B c It represents the number of resource blocks allocated to cell c, and B represents the total number of available resource blocks in the scenario of multi-cell power heterogeneous network.

6. The multi-service coexistence method for multi-cell scenarios of the power Internet of Things according to claim 4 is characterized in that: The calculating the sum of downlink rates of all eMBB users in any cell and constructing a multi-objective optimization problem about the sum of downlink rates of all eMBB users in each cell includes: The downlink rate and f of all eMBB users in cell c c It is expressed as: Then the downlink rate sum of all eMBB users in each cell is expressed as {f1,f2,…,f C }, the multi-objective optimization problem is expressed as:

7. The multi-service coexistence method for multi-cell scenarios of the power Internet of Things according to claim 6 is characterized in that: The NSGA-Ⅲ algorithm is used to solve the multi-objective optimization problem and obtain the Pareto solution set P, {a m,c ,a n,c ,p m,b,c (t),s b (t),x m,b }∈P.

8. The multi-service coexistence method for multi-cell scenarios of the electric power Internet of Things according to claim 7 is characterized in that: The method of obtaining the optimal decision variables regarding user association and resource allocation as a load balancing solution based on the Pareto solution set includes: Based on the NSGA-Ⅲ algorithm, after solving the multi-objective optimization problem and obtaining the Pareto solution set, the following formula is further solved: The optimal decision variable {a m,c ,a n,c ,p m,b,c (t),s b (t),x m,b } as a load balancing solution.

9. A multi-service coexistence device for multi-cell scenarios of power Internet of Things, characterized in that: include: Business scenario building module, decision variable modeling module, optimization problem building module and optimization problem solving module; The service scenario construction module is used to construct a multi-service coexistence implementation mode including eMBB services and uRLLC services in a multi-cell scenario; The decision variable modeling module is used to perform decision variable modeling on user association and resource allocation for coexistence of multiple services in an implementation mode, wherein the decision variables for user association include a base station selection strategy for eMBB users and a base station selection strategy for uRLLC users, and the decision variables for resource allocation include a resource block allocation method, a puncturing density of uRLLC users in different resource blocks, and a base station's transmit power allocation to eMBB users; The optimization problem construction module is used to define the downlink rate obtained by the eMBB user in any cell, the downlink rate obtained by the uRLLC user in any cell, the high reliability and low latency constraints of the uRLLC service constructed based on the downlink rate obtained by the uRLLC user in any cell, the base station power constraints, and the base station resource block constraints according to the decision variable modeling, calculate the downlink rate sum of all eMBB users in any cell, and construct a multi-objective optimization problem about the downlink rate sum of all eMBB users in each cell; The optimization problem solving module is used to solve the Pareto solution set that maximizes the downlink rate of all eMBB users in each cell under constraints, and obtain the optimal decision variables for user association and resource allocation as a load balancing solution based on the Pareto solution set, and realize multi-service coexistence deployment based on the load balancing solution.

10. An electronic device comprising a memory and one or more processors, wherein the memory is used to store a computer program, characterized in that: The processor is used to implement the multi-service coexistence method for the multi-cell scenario of the power Internet of Things as described in any one of claims 1-8 when executing the computer program.

Images