Disclosure of Invention
Embodiments of the present invention provide a method and an apparatus for jointly allocating an IP-optical network communication service, so as to solve the technical problems proposed in the foregoing background art, or at least partially solve the technical problems proposed in the foregoing background art.
In a first aspect, an embodiment of the present invention provides a method for jointly allocating an IP-optical network communication service, where the method includes:
acquiring service information;
inputting the service information into a preset route distribution model, and analyzing the preset route distribution model by using a preset deep reinforcement learning model to obtain service route distribution information;
the preset routing distribution model is obtained by carrying out normalization weighting on the risk balance degree information of the IP-optical network, the similarity degree information of the main and standby routes and the network delay information.
More specifically, before the step of inputting the service information into the preset route distribution model, the method further includes:
acquiring link fault repairing time and fault-free operation time, and acquiring node fault repairing time and fault-free operation time at the same time;
determining link fault probability information according to the link fault repairing time and the fault-free operation time, and determining node fault probability information according to the acquired node fault repairing time and the fault-free operation time;
obtaining node bearing capacity information and link bearing capacity information according to the IP node bearing capacity information and the link bearing capacity information and the optical network node bearing capacity information and the link bearing capacity information;
obtaining link risk information according to the link fault probability information and the link bearing capacity information, obtaining node risk information according to the node bearing capacity information and the node fault probability information, and obtaining IP-optical network risk balance degree information according to the node risk information and the link risk information.
More specifically, before the step of inputting the service information into the preset route distribution model, the method further includes:
acquiring information of a main route link and information of a standby route link to obtain information of an overlapped link of a main route and a standby route;
acquiring information of a main routing node and information of a standby routing node to obtain information of a main routing node and a standby routing node;
and obtaining the similarity information of the main and standby routes according to the overlapped link information and the overlapped node information.
More specifically, before the step of inputting the service information into the preset route distribution model, the method further includes:
acquiring the propagation delay, the processing delay, the queuing delay and the sending delay of a service IP packet in a link;
and obtaining network delay information according to the propagation delay, the processing delay, the queuing delay and the sending delay.
More specifically, before the step of analyzing the preset route distribution model by using the preset deep reinforcement learning model to obtain the service route distribution information, the method further includes:
acquiring random initial service state information, resetting public network gradient information, and synchronizing the public network gradient information into each thread neural network;
obtaining action information and reward information according to the random initial service state information and by combining an action selection strategy, and updating the service state information;
and performing thread training step number self-increment and network iteration time self-increment, calculating the objective function value of each thread when the thread training step number self-increment reaches a preset step number, so as to update the common network gradient information, and synchronizing the updated common network gradient information into each thread to obtain a preset routing distribution model.
More specifically, the preset route distribution model specifically includes:
St.
Δ
ij=|τ
i-τ
j|≤δ,
j∈V,(i,j)∈E
tln is network delay information, BA (t) is IP-optical network risk balance degree information, WijFor link eijMaximum bandwidth, ΔijIs the voltage level difference of the adjacent sites.
In a second aspect, an embodiment of the present invention provides an IP-optical network communication service joint allocation apparatus, including:
the acquisition module is used for acquiring the service information;
the distribution module is used for inputting the service information into a preset route distribution model, and analyzing the preset route distribution model by using a preset deep reinforcement learning model to obtain service route distribution information;
the preset routing distribution model is obtained by carrying out normalization weighting on the risk balance degree information of the IP-optical network, the similarity degree information of the main and standby routes and the network delay information.
In a third aspect, an embodiment of the present invention provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the method for joint allocation of IP-optical network communication services according to the first aspect when executing the program.
In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the IP-optical network communication service joint allocation method according to the first aspect.
According to the IP-optical network communication service joint distribution method and device provided by the embodiment of the invention, the route distribution index is constructed by considering two aspects of the efficiency and the risk balance of the intelligent power grid communication network, the safety problem of the power communication network is not considered from the risk balance aspect singly, the working efficiency problem of the power communication network is not considered from the service delay aspect only, and the influence on the two aspects is considered. The specific evaluation method of the service balance degree and the transmission delay in the power communication network is analyzed, then the weighted analysis calculation is carried out on the service balance degree and the transmission delay, a reasonable preset route distribution model is finally designed, a preset deep reinforcement learning model is selected, and the optimal path is solved, so that the efficient and reasonable operation of the IP-optical network is guaranteed.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a schematic flow chart of a joint allocation method for IP-optical network communication services described in an embodiment of the present invention, as shown in fig. 1, including:
step S1, acquiring service information;
step S2, inputting the service information into a preset route distribution model, and analyzing the preset route distribution model by using a preset deep reinforcement learning model to obtain service route distribution information;
the preset routing distribution model is obtained by carrying out normalization weighting on the risk balance degree information of the IP-optical network, the similarity degree information of the main and standby routes and the network delay information.
Specifically, the preset route allocation module described in the embodiment of the present invention is configured to determine the optimal service route allocation of the service information when the service information is obtained.
The preset routing distribution model described in the embodiment of the invention considers the network delay information and the IP-optical network risk balance degree information of the service and simultaneously considers the similarity information of the main and standby routes, and performs weighting calculation to determine the preset routing distribution model.
The preset routing distribution model described in the embodiment of the present invention specifically includes:
St.
wherein, Tl nBA (t) is network delay information, BA (t) is IP-optical network risk balance degree information, α and β are weighting coefficients of network delay and network risk balance degree respectively, WijFor link eijMaximum bandwidth, ΔijIs the voltage level difference of the adjacent sites.
The analyzing of the preset route distribution model by the preset deep reinforcement learning model described in the embodiment of the invention means that the preset route distribution model is solved according to the preset deep reinforcement learning model so as to obtain ideal service route distribution information, and then the ideal service route distribution information is distributed according to the main route and the standby route respectively.
In the embodiment of the invention, the service information is obtained in batches in a practical application scene, so that the service information is often simultaneously obtained, and therefore, the service information is often required to be simultaneously distributed with routes, and all paths which meet the constraint are solved for each service by using a K shortest path algorithm based on Dijkstra in advance.
After acquiring the random initial service state information, the preset deep reinforcement learning model described in the embodiment of the invention is input into the central network, and the central network in the reinforcement learning model achieves the convergence effect through a certain number of asynchronous training, so that the preset deep reinforcement learning model is obtained.
The embodiment of the invention constructs the route distribution index by considering two aspects of the efficiency and the risk balance of the smart grid communication network, does not consider the safety problem of the power communication network from the risk balance aspect singly, does not consider the working efficiency problem of the power communication network from the service delay aspect, and considers the influence on the two aspects. The specific evaluation method of the service balance degree and the transmission delay in the power communication network is analyzed, then the weighted analysis calculation is carried out on the specific evaluation method and the specific evaluation method, a reasonable preset route distribution model is finally designed, a deep reinforcement learning route planning method is selected, and the optimal path is solved, so that the efficient and reasonable operation of the IP-optical network is guaranteed.
On the basis of the foregoing embodiment, before the step of inputting the service information into a preset route distribution model, the method further includes:
acquiring link fault repairing time and fault-free operation time, and acquiring node fault repairing time and fault-free operation time at the same time;
determining link fault probability information according to the link fault repairing time and the fault-free operation time, and determining node fault probability information according to the acquired node fault repairing time and the fault-free operation time;
obtaining node bearing capacity information and link bearing capacity information according to the IP node bearing capacity information and the link bearing capacity information and the optical network node bearing capacity information and the link bearing capacity information;
obtaining link risk information according to the link fault probability information and the link bearing capacity information, obtaining node risk information according to the node bearing capacity information and the node fault probability information, and obtaining IP-optical network risk balance degree information according to the node risk information and the link risk information.
Specifically, in an IP network, the bandwidth requirement of a service often directly affects the processing capability of a route, and it is necessary to allocate corresponding bearable traffic to different node routes according to their importance. In optical networks, it is desirable to distribute the amount of traffic as evenly as possible.
The risk and the failure probability of the links and the nodes described in the embodiment of the invention are in direct proportion and in inverse proportion to the bearing capacity. The failure probability is measured by the failure times of the nodes and the links and the repair time of each failure, which is a parameter changing with time,
and
is defined as follows:
wherein,
and
respectively representing links e
ijThe first time fault repair time and the no fault commissioning run time,
and
respectively represent nodes v
kFault repair time for the l-th time and fault-free commissioning time.
The bearing capacity of the routing node is determined by the node importance and the flow bearing capacity of a link connected with the node, and the bearing capacity of the link is determined by the link importance, the bandwidth availability and the bearing capacity of two bearing endpoints. Thus, defining IP layer node bearing capacity
And link bearing capacity
The following were used:
wherein,
and
respectively represent nodes v
iAnd link e
ijM represents a node v
iSet of adjacent nodes of, F
ijAnd S
ijRespectively representing links e
ijResidual bandwidth and total bandwidth of V
ipAnd E
ipRespectively representing a set of IP layer nodes and a set of IP layer links.
In the optical network layer, the bearing capacity of the optical transmission device is determined by the node importance, the number of services carried by the node and the service importance, and the bearing capacity of the optical link is determined by the link importance, the number of services carried by the link and the service importance. Thus, defining optical network node bearing capacity
And link bearing capacity
The following were used:
wherein,
and
respectively represent node vi and link e
ijThe importance of (a) to (b),
and
respectively represent the passing of a node v
iAnd link e
ijAll of the services of (a) are,
and
representation service s
kThe business importance of.
Risk of link
And risk of nodes
The definition is as follows:
wherein,
indicating link e at time t
ijThe probability of failure of (a) is,
indicating the probability of failure of node vk at time t,
represents a link e
ijThe bearing capacity of the steel wire rope is improved,
representing a node v
kThe bearing capacity of (1).
In summary, the network risk balance of the IP-optical network is defined as:
the network risk balance degree is represented by the weighted sum of the variances of the link risk and the node risk, and m and n respectively represent the number of links and nodes in the network.
On the basis of the foregoing embodiment, before the step of inputting the service information into a preset route distribution model, the method further includes:
acquiring information of a main route link and information of a standby route link to obtain information of an overlapped link of a main route and a standby route;
acquiring information of a main routing node and information of a standby routing node to obtain information of a main routing node and a standby routing node;
and obtaining the similarity information of the main and standby routes according to the overlapped link information and the overlapped node information.
Specifically, in order to ensure safe and reliable operation of the power communication network, a reasonable standby route needs to be allocated to each service while route allocation is performed to each service. When a node or a link fails, the routing of the service is switched in time, and the reliability of service transmission is ensured. Therefore, in order to reduce the risk after the route switching, the overlapping degree of the links and nodes between the active route and the standby route needs to be as low as possible, and when the standby route is allocated to the service, the similarity of the active route and the standby route needs to be consideredDefining master and slave route similarity
The following were used:
wherein,
and
respectively representing services s
kThe number of overlapped links of the main route and the standby route and the total number of the links,
and
respectively representing services s
kThe number of overlapped nodes of the main route and the standby route and the total number of the nodes.
On the basis of the foregoing embodiment, before the step of inputting the service information into a preset route distribution model, the method further includes:
acquiring the propagation delay, the processing delay, the queuing delay and the sending delay of a service IP packet in a link;
and obtaining network delay information according to the propagation delay, the processing delay, the queuing delay and the sending delay.
The network delay information specifically includes:
wherein,
for a service s
nThe first path of (a) is,
and
the variables are 0-1, and the values are as follows:
wherein, t
rRepresenting the propagation delay, t, of the traffic IP packet in the link
sIndicating the transmission delay of the IP traffic,
indicating that the traffic is at node v
kThe processing time delay of (2) is,
indicating that the traffic is at node v
kIncluding queuing latency in the input queue and latency in the output queue awaiting forwarding. For propagation delay and transmission delay, they are calculated as follows:
wherein L isijDenotes the length of the link (i, j), c denotes the propagation speed of the information in the link, n1Is the refractive index of the fiber.
The embodiment of the invention constructs the route distribution index by considering two aspects of the efficiency and the risk balance of the smart grid communication network, does not consider the safety problem of the power communication network from the risk balance aspect singly, does not consider the working efficiency problem of the power communication network from the service delay aspect, and considers the influence on the two aspects. Specific evaluation methods of service balance and transmission delay in the power communication network are analyzed, then weighting analysis calculation is carried out on the specific evaluation methods, and finally a reasonable preset routing distribution model is designed.
On the basis of the foregoing embodiment, before the step of analyzing the preset route distribution model by using a preset deep reinforcement learning model to obtain service route distribution information, the method further includes:
acquiring random initial service state information, resetting public network gradient information, and synchronizing the public network gradient information into each thread neural network;
obtaining action information and reward information according to the random initial service state information and by combining an action selection strategy, and updating the service state information;
and performing thread training step number self-increment and network iteration time self-increment, calculating the objective function value of each thread when the thread training step number self-increment reaches a preset step number, so as to update the common network gradient information, and synchronizing the updated common network gradient information into each thread to obtain a preset routing distribution model.
Specifically, a state entity in the defined algorithm is a representation set of all services, and a state space of the state entity is as follows:
S={p1,p2,p3,…,pl}
wherein l is the number of services, pi(i-1, 2, …, l) represents a path for the ith traffic, and the path space P for each trafficiAll paths in (i ═ 1,2, …, l) are sorted and indexed, and the index of each state component has +1,0, -1 actions, so that the size of the whole action space a is 3lI.e. for each state S, there are 3lAn optional action is available for selection.
In the process of learning each thread, the selection of the action a adopts the following strategy pi (a)k|sk(ii) a θ'): an action is randomly chosen with a probability of ∈ and the largest estimate V(s) is chosen with a probability of 1- ∈k;θk) The corresponding action value epsilon is gradually reduced along with the increase of the training times.
In the process of learning each thread, the selection of the action a adopts the following strategy pi (a)k|sk(ii) a θ'): an action is randomly chosen with a probability of ∈ and the largest estimate V(s) is chosen with a probability of 1- ∈k;θk) The corresponding action value epsilon is gradually reduced along with the increase of the training times.
Fig. 2 is a flowchart of a preset routing allocation model algorithm according to an embodiment of the present invention, as shown in fig. 2,
step 201, initializing the number of iterations of the entire network, where K is 1;
step 202, initializing the training step number in each thread, wherein k is 1, and step 203, randomly initializing the service state sk of the IP service;
step 204, resetting the gradient d theta of the public network to 0 and
d theta v0 and synchronizing parameters in the public network into the neural network of each thread, θ ═ θ, θ'
v=θ
v(ii) a
Step 205, according to the strategy pi (a)
k|s
k(ii) a θ') is a state selection action, resulting in a reward and a new state;
step 206, step number is increased, K ← K + 1; if k does not reach the preset maximum number of steps in
step 207, return to step 205 to continue the selection action;
step 208, calculate the objective function for each
threadStep 209, for each state
The gradients d θ and d θ are cumulatively updated as follows
v;
R←rk+γR
Step 210, using the gradients do and dovAsynchronous deviceNew theta and thetav(ii) a If K does not reach the preset maximum iteration number, the step 211 returns to the step 202 to continue execution; otherwise, training is finished. If K does not reach the preset maximum iteration number, step 212 is entered, an initial state is given randomly, the initial state is input into the central network, iteration is carried out for a certain number of times, and a routing distribution result is obtained.
The embodiment of the invention constructs the route distribution index by considering two aspects of the efficiency and the risk balance of the smart grid communication network, does not consider the safety problem of the power communication network from the risk balance aspect singly, does not consider the working efficiency problem of the power communication network from the service delay aspect, and considers the influence on the two aspects. The specific evaluation method of the service balance degree and the transmission delay in the power communication network is analyzed, then the weighted analysis calculation is carried out on the specific evaluation method and the specific evaluation method, a reasonable preset route distribution model is finally designed, a deep reinforcement learning route planning method is selected, and the optimal path is solved, so that the efficient and reasonable operation of the IP-optical network is guaranteed.
According to the embodiment of the invention, the preset route distribution model is solved through the preset route distribution model, and an optimal route scheme is found to ensure the healthy and efficient operation of the network, so that the overall time delay can be ensured, the overall risk balance of the network can be ensured, and the healthy, safe, reliable and efficient operation of the IP-optical intelligent power grid communication network is ensured.
In another embodiment of the invention, in order to verify the feasibility and the effectiveness of service route allocation in the scenario of an IP-optical power communication network based on SDN control, an IP-optical power communication network topology in a certain area is selected and simulation analysis is performed. By utilizing training and verification of the A3C algorithm, appropriate primary and backup routes are allocated for IP dataclass services, fig. 3 is a topology structure diagram of an IP-optical power communication network described in an embodiment of the present invention, as shown in fig. 3, the network is composed of IP routing devices and wavelength division multiplexing devices, and an SDN controller performs centralized control on the IP routing devices and provides a programmable interface. The network topology is composed of three types of voltage-class stations, which are divided into 1000kV,750kV and 500kV stations, and are connected through links with corresponding voltage classes. Wherein, the IP layer routing equipment has 8 nodes, the optical layer equipment has 10 nodes, and the network carries 8 services in total. The numbers on the nodes represent the station numbers and the numbers on the links represent the distance and bandwidth between the stations.
The model is trained and solved using the A3C algorithm. Wherein, the A3C neural network structure of the public part corresponds to parameters theta, omega, the A3C neural network structure of the sub-thread corresponds to parameters theta ', omega', the number of globally shared iteration rounds K and the global maximum iteration number KmaxMaximum number of steps k in a single iteration within a threadmaxThe state feature dimension l, the action set A, the reward rt, the discount factor gamma, and the learning rate epsilon.
The specific parameters in the algorithm are set as shown in table 1:
TABLE 1
Firstly, training a network model by using an A3C algorithm, then performing route allocation on 8 preset services by using the model, and obtaining specific data of an IP-optical combined route allocation result, as shown in Table 2:
TABLE 2 deep reinforcement learning algorithm paths corresponding to each service
Fig. 4 is a schematic diagram of a service sequence number 1 route allocation result described in an embodiment of the present invention, and as shown in fig. 4, a solid arrow indicates an allocation result of a primary route, and a dashed arrow indicates an allocation result of a standby route.
For the A3C model, the number of steps is set to 1000 in the interior of a single thread, the data generated in each 1000 states is a group, and the gradients d theta and d theta are cumulatively calculatedvThen using the cumulative gradients do and dovAsynchronously updating theta and thetav. Iteratively training update parameters theta and thetavAnd obtaining the distribution models of the main route and the standby route respectively 1000 times. Inputting preset 8 services into the network for iterative test, randomly setting the initial state for multiple timesAfter the test, a better convergence effect can be obtained. In order to prevent obtaining a local optimal solution, a group with the minimum objective function after convergence in multiple test results is taken, and the solution is considered as a global optimal solution, fig. 5 is a schematic diagram of an optimal solution obtaining process in a random initial state according to an embodiment of the present invention, as shown in fig. 5, in this initial state, the objective function of the primary route test result tends to converge after about 200 iterations, the objective function of the backup route test result tends to converge after about 250 iterations, and the objective function values of the objective functions are 5.19 and 7.98, respectively. At this time, a joint route allocation scheme of the IP-optical network is obtained.
Fig. 6 is a graph of results obtained by comparing network delay, risk balance and overall target of the three algorithms described in the embodiment of the present invention, as shown in fig. 6, the A3C algorithm obtains the best effect in the distribution of the primary route, and compared with the ant colony algorithm, the network average delay, risk balance and overall target are respectively increased by 2.8%, 1.6% and 2.6%; compared with a simulated annealing algorithm, the average time delay of the network, the risk balance and the overall target are respectively improved by 11.0%, 5.8% and 8.3%. Compared with the ant colony algorithm, the three indexes are not changed greatly in the aspect of allocating the standby route; compared with a simulated annealing algorithm, the average time delay of the network, the risk balance degree and the overall target are respectively improved by 15.1%, 6.7% and 11.4%.
Compared with the simulated annealing algorithm, the method has the advantage that the effect of the algorithm is obviously improved, because the simulated annealing algorithm is easy to fall into local optimization. Compared with the ant colony algorithm, the effect is not obviously improved, and the spare route distribution result is even slightly lower than that of the ant colony algorithm. The greatest advantage of A3C is that the convergence rate of the model is improved significantly when the state scale is large.
In summary, the deep reinforcement learning algorithm A3C has a better or similar effect compared with the existing algorithm, but the efficiency is greatly improved.
Fig. 7 is a schematic structural diagram of a joint IP-optical network communication service distribution device according to an embodiment of the present invention, as shown in fig. 7, including: an acquisition module 710 and a distribution module 720; the obtaining module 710 is configured to obtain service information; the distribution module 720 is configured to input the service information into a preset route distribution model, and analyze the preset route distribution model by using a preset deep reinforcement learning model to obtain service route distribution information; the preset routing distribution model is obtained by carrying out normalization weighting on the risk balance degree information of the IP-optical network, the similarity degree information of the main and standby routes and the network delay information.
The apparatus provided in the embodiment of the present invention is used for executing the above method embodiments, and for details of the process and the details, reference is made to the above embodiments, which are not described herein again.
The embodiment of the invention constructs the route distribution index by considering two aspects of the efficiency and the risk balance of the smart grid communication network, does not consider the safety problem of the power communication network from the risk balance aspect singly, does not consider the working efficiency problem of the power communication network from the service delay aspect, and considers the influence on the two aspects. The specific evaluation method of the service balance degree and the transmission delay in the power communication network is analyzed, then the weighted analysis calculation is carried out on the service balance degree and the transmission delay, a reasonable preset route distribution model is finally designed, a preset deep reinforcement learning model is selected, and the optimal path is solved, so that the efficient and reasonable operation of the IP-optical network is guaranteed.
Fig. 8 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 8, the electronic device may include: a processor (processor)810, a communication Interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication Interface 820 and the memory 830 communicate with each other via the communication bus 840. The processor 810 may call logic instructions in the memory 830 to perform the following method: acquiring service information; inputting the service information into a preset route distribution model, and analyzing the preset route distribution model by using a preset deep reinforcement learning model to obtain service route distribution information; the preset routing distribution model is obtained by carrying out normalization weighting on the risk balance degree information of the IP-optical network, the similarity degree information of the main and standby routes and the network delay information.
In addition, the logic instructions in the memory 830 may be implemented in software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
An embodiment of the present invention discloses a computer program product, which includes a computer program stored on a non-transitory computer readable storage medium, the computer program including program instructions, when the program instructions are executed by a computer, the computer can execute the methods provided by the above method embodiments, for example, the method includes: acquiring service information; inputting the service information into a preset route distribution model, and analyzing the preset route distribution model by using a preset deep reinforcement learning model to obtain service route distribution information; the preset routing distribution model is obtained by carrying out normalization weighting on the risk balance degree information of the IP-optical network, the similarity degree information of the main and standby routes and the network delay information.
Embodiments of the present invention provide a non-transitory computer-readable storage medium storing server instructions, where the server instructions cause a computer to execute the method provided in the foregoing embodiments, for example, the method includes: acquiring service information; inputting the service information into a preset route distribution model, and analyzing the preset route distribution model by using a preset deep reinforcement learning model to obtain service route distribution information; the preset routing distribution model is obtained by carrying out normalization weighting on the risk balance degree information of the IP-optical network, the similarity degree information of the main and standby routes and the network delay information.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.