Cloud gaming routing scheduling method, device, storage medium and apparatus

Abstract

The present invention relates to the field of cloud gaming technology, and discloses a cloud gaming routing scheduling method, device, storage medium, and apparatus. The method comprises: obtaining scheduling reference information of candidate scheduling schemes, determining a target scheduling scheme based on the scheduling reference information through a preset cloud gaming routing scheduling model, wherein the preset cloud gaming routing scheduling model is a multi-objective optimization model, and performing cloud gaming routing scheduling according to the target scheduling scheme. Since the present invention introduces a multi-objective optimized preset cloud gaming routing scheduling model to determine the target scheduling scheme, and performs cloud gaming routing scheduling according to the target scheduling scheme, routing scheduling is achieved from multiple dimensions, thereby improving the rationality of routing scheduling.

Description

Technical Field

[0001] The present invention relates to the field of cloud gaming technology, and in particular to a cloud gaming routing scheduling method, device, storage medium, and apparatus. Background Art

[0002] Currently, when performing cloud gaming routing scheduling, the traditional routing scheduling model only considers the influence of some factors, resulting in unreasonable scheduling problems. Summary of the Invention

[0003] The main purpose of the present invention is to provide a cloud gaming routing scheduling method, device, storage medium and apparatus, aiming to solve the technical problem that the traditional routing scheduling model in the prior art only considers the influence of some factors, resulting in unreasonable scheduling.

[0004] To achieve the above objectives, the present invention provides a cloud game routing scheduling method, which includes the following steps:

[0005] Obtaining scheduling reference information of candidate scheduling solutions;

[0006] Determine a target scheduling solution based on the scheduling reference information through a preset cloud game routing scheduling model, wherein the preset cloud game routing scheduling model is a multi-objective optimization model;

[0007] Perform cloud gaming routing scheduling according to the target scheduling scheme.

[0008] Optionally, the scheduling reference information includes the number of edge nodes, the number of users, the number of link segments under the network path, network congestion, network form, throughput, edge node specification information, and game startup hardware conditions;

[0009] The preset cloud game routing scheduling model is:

[0010]

[0011] Where w l , l=1,2,…,6 are the weight coefficients of each part, N is the number of users, M is the number of edge nodes, k = 1, 2, ..., 4 is the scheduling coefficient of each scheduling reference information, indicating whether the i-th user will be scheduled to the j-th edge node when launching the cloud game under the k-th scheduling reference information. When the i-th user starts the cloud game, it is dispatched to the j-th edge node. When the i-th user starts the cloud game, it is not dispatched to the j-th edge node. ij is the number of link segments in the network path between the i-th user and the j-th edge node, Cyb ijis the network congestion between the i-th user and the j-th edge node, W i is the network form of the i-th user, Thr ij is the throughput between the i-th user and the j-th edge node, Spec j is the edge node specification information of the jth edge node, H i Hardware conditions for starting the game for user i, is the variance of the number of idle instances of each edge node, N * The number of queues for users.

[0012] Optionally, the step of obtaining scheduling reference information of the candidate scheduling scheme includes:

[0013] Obtain a list of edge nodes and select the user to be scheduled from the candidate scheduling schemes;

[0014] Determining network information between the user to be scheduled and each edge node in the edge node list;

[0015] Scheduling reference information of the candidate scheduling solution is generated according to the network information.

[0016] Optionally, before the step of obtaining the scheduling reference information of the candidate scheduling solution, the step further includes:

[0017] Generate a target routing scheduling set based on user information and a preset heuristic pseudo-random strategy;

[0018] A candidate scheduling solution is generated according to the target routing scheduling set.

[0019] Optionally, the step of generating a target routing scheduling set according to user information and a preset heuristic pseudo-random strategy includes:

[0020] Generate an initial routing scheduling set based on user information and a preset heuristic pseudo-random strategy;

[0021] Selecting an excellent routing scheduling set from the initial routing scheduling set based on a preset heuristic selection strategy;

[0022] Performing local random mutation on the excellent routing scheduling set with an adaptive probability to obtain a mutated routing scheduling set;

[0023] A target routing scheduling set is selected from the mutated routing scheduling set.

[0024] Optionally, the step of selecting a target routing scheduling set from the mutated routing scheduling set includes:

[0025] Obtaining the mean and variance of each dimension in the mutated routing scheduling set;

[0026] Calculating the normal distribution probability of the mutated routing scheduling set according to the mean and the variance through a preset normal distribution probability model;

[0027] A target routing scheduling set is selected from the mutated routing scheduling set based on the normal distribution probability.

[0028] Optionally, the step of generating a target routing scheduling set according to user information and a preset heuristic pseudo-random strategy includes:

[0029] extracting the user's level information, location information, and recent routing information from the user information;

[0030] A target routing scheduling set is generated according to the level information, the location information and the recent routing information through a preset heuristic pseudo-random strategy.

[0031] In addition, to achieve the above-mentioned purpose, the present invention also proposes a cloud game routing scheduling device, which includes a memory, a processor, and a cloud game routing scheduling program stored on the memory and runnable on the processor, and the cloud game routing scheduling program is configured to implement the cloud game routing scheduling method described above.

[0032] In addition, to achieve the above-mentioned purpose, the present invention also proposes a storage medium, on which a cloud game routing scheduling program is stored. When the cloud game routing scheduling program is executed by a processor, the cloud game routing scheduling method described above is implemented.

[0033] In addition, to achieve the above-mentioned purpose, the present invention also proposes a cloud game routing scheduling device, which includes: an information acquisition module, a solution determination module, and a routing scheduling module;

[0034] The information acquisition module is used to obtain scheduling reference information of candidate scheduling solutions;

[0035] The solution determination module is used to determine a target scheduling solution based on the scheduling reference information through a preset cloud game routing scheduling model, where the preset cloud game routing scheduling model is a multi-objective optimization model;

[0036] The routing scheduling module is used to perform cloud game routing scheduling according to the target scheduling plan.

[0037] In the present invention, the scheduling reference information for obtaining candidate scheduling schemes is disclosed, and the target scheduling scheme is determined through a preset cloud game routing scheduling model based on the scheduling reference information. The preset cloud game routing scheduling model is a multi-objective optimization model, and cloud game routing scheduling is performed according to the target scheduling scheme. Since the present invention introduces a multi-objective optimized preset cloud game routing scheduling model to determine the target scheduling scheme, and performs cloud game routing scheduling according to the target scheduling scheme, routing scheduling is realized from multiple dimensions, thereby improving the rationality of routing scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a schematic diagram of the structure of a cloud gaming routing and scheduling device in a hardware operating environment according to an embodiment of the present invention;

[0039] Figure 2 This is a flow chart of the first embodiment of the cloud gaming routing scheduling method of the present invention;

[0040] Figure 3 This is a flow chart of the second embodiment of the cloud gaming routing scheduling method of the present invention;

[0041] Figure 4 This is a flow chart of the third embodiment of the cloud gaming routing and scheduling method of the present invention;

[0042] Figure 5 This is a structural block diagram of the first embodiment of the cloud gaming routing and scheduling device of the present invention.

[0043] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0044] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0045] Reference Figure 1 , Figure 1 This is a schematic diagram of the structure of a cloud gaming routing and scheduling device in the hardware operating environment involved in an embodiment of the present invention.

[0046] like Figure 1As shown, the cloud gaming routing and scheduling device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), and the user interface 1003 may optionally include a standard wired interface and a wireless interface. The wired interface of the user interface 1003 may be a USB interface in the present invention. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity (Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable memory (NVM), such as a disk memory. The memory 1005 may optionally be a storage device independent of the aforementioned processor 1001.

[0047] Those skilled in the art will understand that Figure 1 The structure shown in does not constitute a limitation on the cloud gaming routing scheduling device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0048] like Figure 1 As shown, the memory 1005 identified as a computer storage medium may include an operating system, a network communication module, a user interface module, and a cloud game routing scheduler.

[0049] exist Figure 1 In the cloud game routing scheduling device shown, the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; the user interface 1003 is mainly used to connect to the user device; the cloud game routing scheduling device calls the cloud game routing scheduling program stored in the memory 1005 through the processor 1001, and executes the cloud game routing scheduling method provided by the embodiment of the present invention.

[0050] Based on the above hardware structure, an embodiment of the cloud game routing scheduling method of the present invention is proposed.

[0051] Reference Figure 2 , Figure 2 This is a flow chart of the first embodiment of the cloud game routing scheduling method of the present invention, and proposes the first embodiment of the cloud game routing scheduling method of the present invention.

[0052] In a first embodiment, the cloud gaming routing scheduling method includes the following steps:

[0053] Step S10: Obtain scheduling reference information of candidate scheduling solutions.

[0054] It should be understood that the execution subject of the method of this embodiment can be a cloud game routing scheduling device with data processing, network communication and program running functions, such as a server, etc., or other electronic devices that can achieve the same or similar functions. This embodiment does not limit this.

[0055] It should be noted that the candidate scheduling schemes may be preset or determined in real time according to user information, and this embodiment does not impose any limitation on this.

[0056] Scheduling reference information may be information that affects cloud gaming routing scheduling, such as the number of edge nodes, the number of users, the number of link segments under the network path, network congestion, network form, throughput, edge node specification information, game startup hardware conditions, and other information.

[0057] It is understandable that obtaining scheduling reference information of a candidate scheduling scheme may be obtaining the users and edge nodes to be scheduled in the candidate scheduling scheme, and obtaining the user information and edge node information to be scheduled, and using the user information and edge node information to be scheduled as scheduling reference information.

[0058] Step S20: Determine a target scheduling solution based on the scheduling reference information through a preset cloud game routing scheduling model, where the preset cloud game routing scheduling model is a multi-objective optimization model.

[0059] It should be noted that the preset cloud game routing scheduling model can be pre-set, and the preset cloud game routing scheduling model can be a multi-objective routing scheduling model with scheduling reference information as a constraint condition.

[0060] It should be understood that after inputting the scheduling reference information into the preset cloud game routing scheduling model, the scheduling reference scores corresponding to the candidate scheduling schemes can be obtained, and the candidate scheduling schemes can be sorted from small to large according to the scheduling reference scores, and the target scheduling scheme can be selected according to the sorting results.

[0061] It is understandable that selecting the target scheduling solution according to the ranking result may be to use the candidate scheduling solution with the highest ranking as the target scheduling solution.

[0062] Step S30: Perform cloud game routing scheduling according to the target scheduling plan.

[0063] It should be understood that performing cloud gaming routing scheduling according to a target scheduling scheme may be scheduling users to corresponding target edge nodes according to the target scheduling scheme.

[0064] In the first embodiment, the scheduling reference information for obtaining candidate scheduling schemes is disclosed, and the target scheduling scheme is determined through a preset cloud game routing scheduling model based on the scheduling reference information. The preset cloud game routing scheduling model is a multi-objective optimization model, and cloud game routing scheduling is performed according to the target scheduling scheme. Since this embodiment introduces a multi-objective optimized preset cloud game routing scheduling model to determine the target scheduling scheme, and performs cloud game routing scheduling according to the target scheduling scheme, routing scheduling is realized from multiple dimensions, thereby improving the rationality of routing scheduling.

[0065] Reference Figure 3 , Figure 3 This is a flow chart of the second embodiment of the cloud game routing scheduling method of the present invention, based on the above Figure 2 The first embodiment shown proposes a second embodiment of the cloud gaming routing scheduling method of the present invention.

[0066] In the second embodiment, step S10 includes:

[0067] Step S101: Obtain a list of edge nodes and select a user to be scheduled from candidate scheduling solutions.

[0068] Step S102: Determine network information between the user to be scheduled and each edge node in the edge node list.

[0069] Step S103: Generate scheduling reference information of the candidate scheduling solution according to the network information.

[0070] It should be understood that in order to improve the accuracy of the scheduling reference information, in this embodiment, the scheduling reference information may be generated based on network information between the user to be scheduled and each edge node in the edge node list.

[0071] For ease of understanding, the following examples are provided, but are not intended to limit this solution. Scheduling reference information is generated through the following steps:

[0072] (1) Initialize parameters, add each edge node to the edge node list, and set the scheduling list of each edge node And the queue list of users at each edge node

[0073] (2) Sequentially select a user (i.e., the user to be scheduled) from the scheduling plan and obtain the user's corresponding information: the user's network address, user level, the network form the user plays, and the user's historical routing information;

[0074] (2.1) Determine whether the user's historical routing information is empty. If it is empty:

[0075] (a) Sequentially select an edge node from the edge node list. If the edge node list has been traversed and the user cannot match a suitable edge node, then a random edge node is selected from the edge list and dispatched to the user, and the user is added to the edge node user queue list. Go to step (e);

[0076] (b) Calculate the network congestion between the user and the selected edge node. If the network congestion between the user and the selected edge node reaches a set threshold, go to step (a);

[0077] (c) Calculate the network throughput between the user and the selected edge node. If the throughput exceeds the network throughput threshold, go to step (a).

[0078] (d) Determine whether all instances of the edge node are occupied, and if so, go to step (a);

[0079] (e) determining whether the rule information of the edge node meets the requirements for the user to start the game, if not, go to step (a);

[0080] (f) Calculate the number of links from this user to the selected edge node in the network path and go to step (2.3);

[0081] (2.2) If the user's historical routing information is not empty, perform routing scheduling condition judgment and data processing according to steps (b) to (f); if the conditions are not met, go to step (a);

[0082] (2.3) Add this user to the edge node scheduling list If all users in the scheduling scheme have been assigned corresponding routing information, go to step (3); otherwise, go to step (2);

[0083] (3) Scheduling reference information based on the network congestion situation, number of link segments, network throughput, number of queued users, etc. of each user in the above steps.

[0084] In the second embodiment, a method is disclosed for obtaining an edge node list, selecting a user to be scheduled from candidate scheduling solutions, determining network information between the user to be scheduled and each edge node in the edge node list, and generating scheduling reference information for the candidate scheduling solution based on the network information. Because the scheduling reference information is generated based on the network information between the user to be scheduled and each edge node in the edge node list in this embodiment, the accuracy of the scheduling reference information can be improved, thereby improving the accuracy of cloud gaming routing scheduling.

[0085] In a second embodiment, the scheduling reference information includes the number of edge nodes, the number of users, the number of link segments under the network path, network congestion, network form, throughput, edge node specification information, and game startup hardware conditions;

[0086] The preset cloud game routing scheduling model is:

[0087]

[0088] Where w l , l=1,2,…,6 are the weight coefficients of each part, N is the number of users, M is the number of edge nodes, k = 1, 2, ..., 4 is the scheduling coefficient of each scheduling reference information, indicating whether the i-th user will be scheduled to the j-th edge node when launching the cloud game under the k-th scheduling reference information. When the i-th user starts the cloud game, it is dispatched to the j-th edge node. When the i-th user starts the cloud game, it is not dispatched to the j-th edge node. ij is the number of link segments in the network path between the i-th user and the j-th edge node, Cyb ij is the network congestion between the i-th user and the j-th edge node, W i is the network form of the i-th user, Thr ij is the throughput between the i-th user and the j-th edge node, Spec j is the edge node specification information of the jth edge node, H i Hardware conditions for starting the game for user i, is the variance of the number of idle instances of each edge node, N * The number of queues for users.

[0089] For ease of understanding, the following examples are provided, but are not intended to limit this solution. The process of establishing a cloud gaming routing and scheduling model specifically includes the following steps:

[0090] Step 1: Obtain the specifications of each edge node and its corresponding instance number, the number of link segments under the network path, network congestion, throughput, as well as the number of users, level information, the network mode in which users play, the user's geographic location and the user's recently active routing information, the hardware conditions for game startup, and perform data preprocessing.

[0091] The specific process of data preprocessing is as follows: 1) Describe the specification information of each edge node in a hierarchical manner and mark them from low to high with 1, 2, and 3, and map them to the hardware information required for the game, with a one-to-one correspondence; 2) Map the number of instances of each specification, the number of link segments under the network path, network congestion, network throughput, user level information, and the network form in which the user plays to form a dimensionless map; 3) Obtain the user's network address based on the user's geographic location.

[0092] Step 2: Establish a cloud game routing scheduling model based on the preprocessed information and transform the model into a quasi-traveling salesman problem model.

[0093] 1) Given the number of users N and the number of edge nodes M, set the specification information of each edge node Spec j and its corresponding number of instances Number of links in the network path Link ij 、Network congestion Cyb ij and throughput Thr ij ; Set the hardware conditions for game startup H i , user level information G i , network form W i , where i = 1, 2, 3,…, N, j = 1, 2, 3,…, M.

[0094] 2) Set the constraints of the cloud gaming routing model:

[0095] (1) Each edge node has the same specification information. The game startup requirements cannot be higher than the specifications of each edge node. Game startup requirements and specification information are prioritized for allocation at the same level.

[0096] (2) Each user can only be dispatched to one edge node when trying to play the game;

[0097] (3) Under the same circumstances, the principle of priority allocation should be followed for users with higher user levels;

[0098] (4) Generally speaking, network throughput is positively correlated with the user's network form. When the network is congested or the throughput reaches a predetermined value, users need to wait in line.

[0099] 3) Based on the above data information and constraints, the cloud gaming routing scheduling model is established as follows:

[0100]

[0101] Where w l , l=1,2,…,6 are the weight coefficients of each part, N is the number of users, M is the number of edge nodes, k = 1, 2, ..., 4 is the scheduling coefficient of each scheduling reference information, indicating whether the i-th user will be scheduled to the j-th edge node when launching the cloud game under the k-th scheduling reference information. When the i-th user starts the cloud game, it is dispatched to the j-th edge node. When the i-th user starts the cloud game, it is not dispatched to the j-th edge node. ij is the number of link segments in the network path between the i-th user and the j-th edge node, Cyb ij is the network congestion between the i-th user and the j-th edge node, W i is the network form of the i-th user, Thr ij is the throughput between the i-th user and the j-th edge node, Spec j is the edge node specification information of the jth edge node, H i Hardware conditions for starting the game for user i, is the variance of the number of idle instances of each edge node, N * The number of users queued.

[0102] in, is calculated as follows:

[0103]

[0104] Indicates whether the i-th user will be scheduled to the j-th edge node when launching the cloud game under the k-th scheduling reference information;

[0105] in, is calculated as follows:

[0106]

[0107] Where, represents the number of idle instances of the jth edge node, Indicates the average number of idle instances of M edge nodes.

[0108] The cloud game routing scheduling model has six parts, namely, the penalty for the number of link segments in the network path when the user is scheduled to each edge node, the network congestion penalty, the throughput penalty, the penalty for the mismatch between the hardware conditions required for game startup and the edge node specification information, the variance penalty for the number of idle instances of each edge node, and the penalty for the number of queued users.

[0109] In the second embodiment, the cloud game routing scheduling model comprehensively considers multiple factors such as the number of edge nodes, the number of users, the number of link segments under the network path, network congestion, throughput, edge node specification information, game startup hardware conditions, etc., which is conducive to improving the instance utilization rate of each edge node, reducing resource waste, and optimizing cloud game routing scheduling problems, reducing the occurrence of phenomena such as users queuing for game trials, slow game startup, and network congestion, thereby improving user experience satisfaction.

[0110] Reference Figure 4 , Figure 4 This is a flow chart of the third embodiment of the cloud game routing scheduling method of the present invention. Based on the above embodiments, the third embodiment of the cloud game routing scheduling method of the present invention is proposed.

[0111] In the third embodiment, before step S10, the method further includes:

[0112] Step S01: Generate a target routing scheduling set according to user information and a preset heuristic pseudo-random strategy.

[0113] It should be understood that traditional intelligent scheduling optimization algorithms mainly include genetic algorithms, simulated annealing algorithms, and ant colony algorithms. Among them, the genetic algorithm has a cumbersome decoding process for scheduling problems, is prone to premature convergence, and has weak optimization capabilities; the simulated annealing algorithm has a simple calculation and algorithm flow, but the algorithm has poor optimization accuracy; the ant colony algorithm has complex pheromone calculations and a cumbersome algorithm flow, making it unsuitable for scheduling problems with high real-time requirements such as cloud gaming routing scheduling; the discrete distribution estimation algorithm, as a new type of randomized optimization algorithm based on statistical principles, has a simple and easy algorithm flow and can be well applied to some problems with high real-time requirements, but the algorithm itself has shortcomings such as insufficient inspiration and a tendency to fall into local optimality.

[0114] Therefore, in order to optimize the large number of infeasible solutions generated by general random strategies, improve the optimization ability of the discrete distribution estimation algorithm, and accelerate the convergence speed of the algorithm, this step adopts a preset heuristic pseudo-random strategy based on user information, which effectively increases the number of feasible solutions in the early stage of the algorithm and improves the optimization speed of the algorithm.

[0115] In a specific implementation, for example, the population is initialized and a heuristic pseudo-random strategy based on the user's recently active routing information (ie, user information) is used to generate a Pop routing scheduling set D of size t (Scheduling scheme), each scheme contains N users, where the number of iterations t is set to 0 at the initial moment and the maximum number of iterations is T.

[0116] Furthermore, in order to improve the reliability of the routing scheduling set, step S01 includes:

[0117] extracting the user's level information, location information, and recent routing information from the user information;

[0118] A target routing scheduling set is generated according to the level information, the location information and the recent routing information through a preset heuristic pseudo-random strategy.

[0119] Furthermore, in order to increase population diversity and improve the global search capability of the algorithm, step S01 further includes:

[0120] Generate an initial routing scheduling set based on user information and a preset heuristic pseudo-random strategy;

[0121] Selecting an excellent routing scheduling set from the initial routing scheduling set based on a preset heuristic selection strategy;

[0122] Performing local random mutation on the excellent routing scheduling set with an adaptive probability to obtain a mutated routing scheduling set;

[0123] A target routing scheduling set is selected from the mutated routing scheduling set.

[0124] It is understandable that in order to prevent the algorithm from falling into a local optimum and enhance the ability of the disadvantaged group and the advantaged group to learn from each other, selecting the target routing scheduling set from the mutated routing scheduling set can be to obtain the mean and variance of each dimension in the mutated routing scheduling set, calculate the normal distribution probability of the mutated routing scheduling set based on the mean and variance through a preset normal distribution probability model, and select the target routing scheduling set from the mutated routing scheduling set based on the normal distribution probability.

[0125] For ease of understanding, the following examples are provided, but are not intended to limit this solution. The steps for generating the target route scheduling set are as follows:

[0126] 1. Initialize the population and use a heuristic pseudo-random strategy based on the user's recently active routing information (i.e. user information) to generate a Pop routing scheduling set D of size t (Scheduling scheme), each scheme contains N users, where the number of iterations t is set to 0 at the initial moment and the maximum number of iterations is T.

[0127] In order to optimize the large number of infeasible solutions generated by general random strategies, improve the optimization ability of the discrete distribution estimation algorithm, and accelerate the convergence of the algorithm, this step adopts a pseudo-random strategy based on the user's recently active routing information, effectively increasing the number of feasible solutions in the early stage of the algorithm and improving the optimization speed of the algorithm.

[0128] (1) N users are divided into groups according to their level G i Sorting: Users of the same level are randomly arranged according to their serial numbers as an initial routing scheduling scheme;

[0129] (2) obtaining the network address of each user according to the geographical location of each user in the scheme;

[0130] (3) Obtain the user's historical network address and its corresponding edge node information based on the user's recently active routing information;

[0131] (4) Determine whether the user's network address at this time is the same as or close to the user's historical network address. If the network address meets the requirements, the edge node information corresponding to the historical network is dispatched to the user. Otherwise, the routing dispatch information of the user is left empty.

[0132] (5) Repeat the above steps to generate a routing scheduling set D of size Pop t .

[0133] 2. According to the preset cloud game routing scheduling model, a decoding device based on the quasi-traveling salesman problem model is used to calculate the objective function value of each individual in the population, and determine whether the termination condition t>T is met. If so, the individual with the smallest objective function value is selected as the optimal result, and the corresponding user routing scheduling result is output. Otherwise, the algorithm process continues.

[0134] 3. Use adaptive selection strategy to select m excellent individuals from the original population to form the optimal evaluation set

[0135] In order to increase the diversity of the population and improve the global search capability of the algorithm, this step adopts a heuristic selection strategy to adaptively improve the diversity and number of excellent individual selections according to the increase in the number of iterations, thereby increasing the global search capability of the algorithm and the diversity of solutions. The specific process is as follows:

[0136] (1) Determine the number of excellent individuals in each generation m, and the calculation formula is as follows:

[0137]

[0138] Where α is the selection factor, which is generally set to 0.5<α<1, ceil(*) is the upward rounding function, t is the current number of iterations, and T is the maximum number of iterations.

[0139] (2) Sort the objective function values ​​of Pop individuals from small to large, and select the top m excellent individuals as the evaluation set

[0140] (3) Evaluation set Perform local random mutations with adaptive probabilities to prevent the population from falling into local optima.

[0141] Specifically, calculate the total objective function value of all individuals in the evaluation set, and calculate the proportion of each individual in the evaluation set to the total objective function value; generate a random value between 0 and 1 for each individual in the evaluation set in turn; if the random value is not less than the proportion of the individual in the total objective function value, randomly perturb the individual to generate a new individual; add all newly generated individuals to the original evaluation set to form a new evaluation set And update the original evaluation set

[0142] 4. Construct a normal distribution probability model and select the best evaluation set according to the probability model. Generate new individuals by sampling.

[0143] Among them, the normal distribution probability model is:

[0144]

[0145] Where, f(x i ) represents the normal distribution probability of each dimension of the population, μ i represents the mean of each dimension in the population, σ i Represents the variance of each dimension in the population, i = 1, 2,…, N.

[0146] The specific process of generating Pop new scheduling plans through sampling is as follows:

[0147] (4.1) Start by generating a new scheduling plan, set the individual dimension i to 1, and add N users to the unscheduled set List;

[0148] (4.2) Calculate the evaluation set The mean μ of the middle dimension i i and variance σ i , and calculate the normal distribution probability parameter f(x i );

[0149] In order to avoid the algorithm from falling into local optimality and enhance the ability of disadvantaged and advantaged groups to learn from each other, a learning factor β is introduced to calculate the mean μ of each dimension. i and variance σ i , which effectively improves the communication ability between individuals in the population, and is conducive to improving the search and optimization capabilities of the algorithm. The calculation formulas are as follows:

[0150]

[0151]

[0152] (4.3) Randomly generate a random number r between 0 and 1, and choose r greater than f(x i) user set, if the user set is not empty, randomly select a user to join the scheduling plan; otherwise select f(x i ) Add the user closest to r to the scheduling plan, remove the user from the List, and repeat this step until the List set is empty;

[0153] (4.4) If N new scheduling schemes are not generated, go to step (4.1), otherwise the sampling process ends and N new scheduling schemes are output.

[0154] Step S02: generating candidate scheduling solutions according to the target routing scheduling set.

[0155] It can be understood that generating candidate scheduling solutions according to the target route scheduling set may be directly using the scheduling solutions in the target route scheduling set as candidate scheduling solutions.

[0156] In the third embodiment, an improved discrete distribution estimation algorithm is based on a dynamic selection strategy and a heuristic pseudo-random strategy to improve the optimization and search capabilities of the original algorithm, and a decoding device based on the quasi-traveling salesman problem is used to solve the objective function of the improved distribution estimation algorithm, thereby reducing the large number of infeasible solutions generated by the original algorithm and improving the algorithm's solution performance.

[0157] In addition, an embodiment of the present invention further proposes a storage medium, on which a cloud game routing scheduling program is stored. When the cloud game routing scheduling program is executed by a processor, the cloud game routing scheduling method described above is implemented.

[0158] In addition, refer to Figure 5 , an embodiment of the present invention further proposes a cloud game routing scheduling device, the cloud game routing scheduling device comprising: an information acquisition module 10, a solution determination module 20 and a routing scheduling module 30;

[0159] The information acquisition module 10 is used to acquire scheduling reference information of candidate scheduling solutions.

[0160] It should be noted that the candidate scheduling schemes may be preset or determined in real time according to user information, and this embodiment does not impose any limitation on this.

[0161] Scheduling reference information may be information that affects cloud gaming routing scheduling, such as the number of edge nodes, the number of users, the number of link segments under the network path, network congestion, network form, throughput, edge node specification information, game startup hardware conditions, and other information.

[0162] It is understandable that obtaining scheduling reference information of a candidate scheduling scheme may be obtaining the users and edge nodes to be scheduled in the candidate scheduling scheme, and obtaining the user information and edge node information to be scheduled, and using the user information and edge node information to be scheduled as scheduling reference information.

[0163] The solution determination module 20 is used to determine the target scheduling solution based on the scheduling reference information through a preset cloud game routing scheduling model, and the preset cloud game routing scheduling model is a multi-objective optimization model.

[0164] It should be noted that the preset cloud game routing scheduling model can be pre-set, and the preset cloud game routing scheduling model can be a multi-objective routing scheduling model with scheduling reference information as a constraint condition.

[0165] It should be understood that after inputting the scheduling reference information into the preset cloud game routing scheduling model, the scheduling reference scores corresponding to the candidate scheduling schemes can be obtained, and the candidate scheduling schemes can be sorted from small to large according to the scheduling reference scores, and the target scheduling scheme can be selected according to the sorting results.

[0166] It is understandable that selecting the target scheduling solution according to the ranking result may be to use the candidate scheduling solution with the highest ranking as the target scheduling solution.

[0167] The routing scheduling module 30 is used to perform cloud game routing scheduling according to the target scheduling plan.

[0168] It should be understood that performing cloud gaming routing scheduling according to a target scheduling scheme may be scheduling users to corresponding target edge nodes according to the target scheduling scheme.

[0169] In this embodiment, the scheduling reference information for obtaining candidate scheduling schemes is disclosed, and the target scheduling scheme is determined through a preset cloud game routing scheduling model based on the scheduling reference information. The preset cloud game routing scheduling model is a multi-objective optimization model, and cloud game routing scheduling is performed according to the target scheduling scheme. Since this embodiment introduces a multi-objective optimized preset cloud game routing scheduling model to determine the target scheduling scheme, and performs cloud game routing scheduling according to the target scheduling scheme, routing scheduling is realized from multiple dimensions, thereby improving the rationality of routing scheduling.

[0170] Other embodiments or specific implementations of the cloud gaming routing scheduling device of the present invention can refer to the above-mentioned method embodiments and will not be repeated here.

[0171] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or system comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or system. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or system comprising the element.

[0172] The serial numbers of the above embodiments of the present invention are for description only and do not represent the advantages or disadvantages of the embodiments.

[0173] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, or of course by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as a read-only memory image (ROM) / random access memory (RAM), a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0174] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A cloud game routing scheduling method, characterized in that: The cloud game routing scheduling method includes the following steps: Obtaining scheduling reference information of candidate scheduling solutions; Determine a target scheduling solution based on the scheduling reference information through a preset cloud game routing scheduling model, wherein the preset cloud game routing scheduling model is a multi-objective optimization model; Perform cloud gaming routing scheduling according to the target scheduling plan; Before obtaining the scheduling reference information of the candidate scheduling scheme, the method further includes: Generate a target routing scheduling set based on user information and a preset heuristic pseudo-random strategy; generating a candidate scheduling solution according to the target routing scheduling set; The generating of the target routing scheduling set according to the user information and the preset heuristic pseudo-random strategy includes: Generate an initial routing scheduling set based on user information and a preset heuristic pseudo-random strategy; Selecting an excellent routing scheduling set from the initial routing scheduling set based on a preset heuristic selection strategy; Performing local random mutation on the excellent routing scheduling set with an adaptive probability to obtain a mutated routing scheduling set; A target routing scheduling set is selected from the mutated routing scheduling set.

2. The cloud game routing scheduling method according to claim 1, characterized in that: The scheduling reference information includes the number of edge nodes, the number of users, the number of link segments under the network path, network congestion, network form, throughput, edge node specification information, and game startup hardware conditions; The preset cloud game routing scheduling model is: Where w l , l=1,2,…,6 are the weight coefficients of each part, N is the number of users, M is the number of edge nodes, k = 1, 2, ..., 4 is the scheduling coefficient of each scheduling reference information, indicating whether the i-th user will be scheduled to the j-th edge node when launching the cloud game under the k-th scheduling reference information. When the i-th user starts the cloud game, it is dispatched to the j-th edge node. When the i-th user starts the cloud game, it is not dispatched to the j-th edge node. ij is the number of link segments in the network path between the i-th user and the j-th edge node, Cyb ij is the network congestion between the i-th user and the j-th edge node, W i is the network form of the i-th user, Thr ij is the throughput between the i-th user and the j-th edge node, Spec j is the edge node specification information of the jth edge node, H i Hardware conditions for starting the game for user i, is the variance of the number of idle instances of each edge node, N * The number of queues for users.

3. The cloud game routing scheduling method according to claim 1, wherein: The step of obtaining scheduling reference information of candidate scheduling solutions includes: Obtain a list of edge nodes and select the user to be scheduled from the candidate scheduling schemes; Determining network information between the user to be scheduled and each edge node in the edge node list; Scheduling reference information of the candidate scheduling solution is generated according to the network information.

4. The cloud game routing scheduling method according to claim 1, wherein: The step of selecting a target routing scheduling set from the mutated routing scheduling set includes: Obtaining the mean and variance of each dimension in the mutated routing scheduling set; Calculating the normal distribution probability of the mutated routing scheduling set according to the mean and the variance through a preset normal distribution probability model; A target routing scheduling set is selected from the mutated routing scheduling set based on the normal distribution probability.

5. The cloud game routing scheduling method according to claim 1, wherein: The step of generating a target routing scheduling set according to user information and a preset heuristic pseudo-random strategy includes: extracting the user's level information, location information, and recent routing information from the user information; A target routing scheduling set is generated according to the level information, the location information and the recent routing information through a preset heuristic pseudo-random strategy.

6. A cloud game routing and scheduling device, characterized in that: The cloud game routing scheduling device includes: a memory, a processor, and a cloud game routing scheduling program stored in the memory and executable on the processor. When the cloud game routing scheduling program is executed by the processor, the cloud game routing scheduling method according to any one of claims 1 to 5 is implemented.

7. A storage medium, characterized in that: The storage medium stores a cloud gaming routing scheduling program, which, when executed by a processor, implements the cloud gaming routing scheduling method according to any one of claims 1 to 5.

8. A cloud game routing scheduling device, characterized in that: The cloud game routing and scheduling device includes: an information acquisition module, a solution determination module, and a routing scheduling module; The information acquisition module is used to obtain scheduling reference information of candidate scheduling solutions; The solution determination module is used to determine a target scheduling solution based on the scheduling reference information through a preset cloud game routing scheduling model, where the preset cloud game routing scheduling model is a multi-objective optimization model; The routing scheduling module is used to perform cloud game routing scheduling according to the target scheduling plan; Before obtaining the scheduling reference information of the candidate scheduling scheme, the method further includes: Generate a target routing scheduling set based on user information and a preset heuristic pseudo-random strategy; generating a candidate scheduling solution according to the target routing scheduling set; The generating of the target routing scheduling set according to the user information and the preset heuristic pseudo-random strategy includes: Generate an initial routing scheduling set based on user information and a preset heuristic pseudo-random strategy; Selecting an excellent routing scheduling set from the initial routing scheduling set based on a preset heuristic selection strategy; Performing local random mutation on the excellent routing scheduling set with an adaptive probability to obtain a mutated routing scheduling set; A target routing scheduling set is selected from the mutated routing scheduling set.

Images