Dynamic distribution method of 4G network and storage medium

Abstract

The invention discloses a dynamic allocation method of a 4G network and a storage medium, and belongs to the technical field of wireless communication. The method comprises the following steps: collecting a plurality of equipment network parameters in real time under the condition that the terminal equipment is in a network connection state; the equipment network parameters comprise a plurality of internal parameters, the internal parameters of the equipment network parameters are input into the network model for reasoning calculation, and a reasoning calculation result corresponding to each internal parameter is generated; performing dynamic planning based on each reasoning calculation result to obtain an optimal resource allocation strategy sequence; judging whether the current network connection state is an optimal network connection state or not according to the optimal resource allocation strategy sequence, and if the judgment result is that the current network connection state is not optimal, executing network switching; and after network switching is executed, a storage mode is planned, and equipment network parameters, a network model, a reasoning calculation result, dynamic planning and an optimal resource allocation strategy sequence are allocated to a reasonable storage space.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technologies, and particularly to the field of dynamic allocation of 4G communication networks. Background Art

[0002] In a modern 4G network environment, with the popularization of mobile terminal devices and the rapid growth of data traffic, how to effectively allocate network resources and optimize connection strategies has become an important problem to be solved urgently. The network connection methods in the prior art usually rely on fixed strategies and static parameter configurations and lack the ability of dynamic adjustment. This method shows obvious deficiencies in a complex and changeable network environment, mainly reflected in the inability to effectively meet the resource allocation requirements in different network states, resulting in low network connection efficiency and poor user experience.

[0003] Currently, the 4G network connection methods usually judge the quality of the network based on preset parameters such as signal strength and network latency and perform network switching. However, these methods often ignore the influence of multi-dimensional factors such as the geographical location of the terminal device, user behavior, and device network status. Due to the lack of comprehensive consideration of multiple dynamically changing parameters, the prior art is difficult to provide an optimal network resource allocation strategy, resulting in low resource utilization rate, and when the network environment changes, it is unable to adjust the strategy in time to ensure the optimality of the connection.

[0004] In addition, after performing network switching, the traditional network connection methods usually do not save and optimize the previous network models and strategies, which further limits the response ability of the system in future similar scenarios. For frequently occurring network environment changes, the prior art is difficult to provide an effective storage and scheduling mechanism to optimize resource allocation and connection strategies. Summary of the Invention

[0005] The purpose of the embodiments of the present application is to address the deficiencies of the above prior art and provide a dynamic allocation method and storage medium for a 4G network, which can solve the problems that the prior art is difficult to provide an optimal network resource allocation strategy, resulting in low resource utilization rate, and when the network environment changes, it is unable to adjust the strategy in time to ensure the optimality of the connection. By collecting multiple device network parameters in real time and performing reasoning calculus and dynamic programming, an optimal resource allocation strategy sequence is obtained, and whether to perform network switching is judged through the optimal resource allocation strategy sequence. This method can dynamically evaluate the current network connection state and ensure that the terminal device always maintains the optimal network connection state. Compared with the traditional network connection method that only relies on signal strength, this method can better adapt to a complex network environment and reduce network connection problems.

[0006] To solve the above technical problems, the embodiments of the present application are implemented as follows:

[0007] In a first aspect, the present application provides a method for dynamic allocation of a 4G network. The method includes: when the terminal device is in a network connection state, collecting multiple device network parameters in real time; the device network parameters include multiple internal parameters, inputting the internal parameters of the device network parameters into a network model for inference and calculation to generate an inference and calculation result corresponding to each internal parameter; based on each inference and calculation result, performing dynamic programming to obtain an optimal resource allocation strategy sequence; judging whether the current network connection state is in an optimal network connection state according to the optimal resource allocation strategy sequence, and if the judgment result is non-optimal, performing network switching; after performing network switching, planning a storage method to allocate the device network parameters, the network model, the inference and calculation results, the dynamic programming, and the optimal resource allocation strategy sequence to a reasonable storage space.

[0008] Further, the internal parameters include geographical location information, device network status information, and user behavior information, and the geographical location information, the device network status information, and the user behavior information are all collected or read by the terminal device.

[0009] Further, the device network status information includes network latency, and the method for obtaining the network latency is: obtaining the network latency data by sending a test data packet and measuring its round-trip time.

[0010] Further, the device network status information includes network congestion status, and the method for obtaining the network congestion status is: obtaining the network congestion status value by analyzing the ratio of the actual data transmission rate to the theoretical maximum rate.

[0011] Further, the network model is a static Bayesian network, and the inference and calculation process includes:

[0012] Establishing a Bayesian network for each of the device network parameters, where each node represents an internal parameter of the device network parameter, the edge represents the dependency relationship between the nodes, and the nodes include parent nodes and child nodes;

[0013] Constructing a conditional probability table, where the conditional probability table includes the prior probability or conditional probability of each node in different states, the prior probability is expressed as the node probability distribution given the parent node in the conditional probability table, and the conditional probability is expressed as the probability distribution of the child node in different states in the conditional probability ratio;

[0014] Verifying whether the conditional probability table is accurate through independent hypothesis verification, and the independent hypothesis verification uses a chi-square test;

[0015] Calculating the joint probability distribution of the nodes, and the calculation formula of the joint probability distribution is:

[0016]

[0017] Wherein, P(X) is the joint probability distribution of the node, and X i represents the i-th node, i is the index of the node, n is the total number of nodes, and Pa(X i ) is the parent node of node X i , and P(X i |Pa(X i )) is the child node conditional probability expression;

[0018] Calculate the posterior probability according to the joint probability distribution, and the calculation of the posterior probability includes applying the marginalization method to the joint probability distribution to remove the unnecessary node conditional probabilities and convert them into posterior probabilities, and using the posterior probabilities as the result of the inference calculation.

[0019] Furthermore, the dynamic programming includes:

[0020] Define the network resource allocation state, decision, and benefit, where the decision is based on the result of the inference calculation; perform state transition through the state transition equation; use the Bellman equation to recursively calculate the maximum cumulative benefit, thereby obtaining the optimal resource allocation strategy sequence;

[0021] Specifically, it includes:

[0022] State definition, the state definition is divided into state S(t), decision a(t), and benefit R(t,a(t)). The state S(t) represents the network resource allocation situation at time t, the decision represents the resource allocation strategy at time t, the resource allocation strategy is based on the result of the inference calculation, and the benefit R(t,a(t)) represents the immediate benefit brought by selecting decision a(t);

[0023] State transition, the state transition transfers the current state to the next state through the state transition equation, and the state transition equation is:

[0024] S(t + 1) = f(S(t), a(t))

[0025] where: S(t + 1) is the next state, and f is the state transition function;

[0026] Furthermore, among them, V(t,S(t)) is the maximum cumulative benefit at time t, γ is the discount factor, is the optimal decision;

[0027] Obtain the optimal resource allocation strategy sequence according to the maximum cumulative benefit through recursive calculation.

[0028] Further, the storage space includes: a random access unit, a sequential access unit, and a fast access unit. According to the data volume and access frequency, the device network parameters, network model, inference calculation results, dynamic programming, and optimal resource allocation strategy sequence are allocated to the random storage unit, sequential access unit, and fast access unit. In a second aspect, an embodiment of the present application provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of a dynamic allocation method for a 4G network are implemented.

[0029] Further, the storage medium includes NAND Flash, and the NAND Flash is allocated as the sequential access unit within the system.

[0030] Further, the storage medium includes NOR Flash, and the NOR Flash is allocated as the fast access unit and the random access unit within the system.

[0031] Advantages of the present invention: The present invention proposes a dynamic allocation method for a 4G network. By collecting multiple network parameters of a device in real time and using a static Bayesian network model for inference calculation, dynamic programming is then performed based on the inference results to obtain an optimal resource allocation strategy sequence. This method can not only comprehensively consider the influence of multiple internal parameters such as geographical location, device network status, and user behavior, but also automatically adjust the network connection status when the network environment changes, ensuring that the terminal device is always in the optimal network connection state; compared with the traditional network connection method that only relies on signal strength, this method can better adapt to complex network environments and reduce network connection problems.

[0032] In addition, the present invention also introduces a reasonable storage strategy to ensure that network parameters, models, inference results, and dynamic programming strategies can be effectively saved and managed, thereby enhancing the response and adaptability of the system in complex network environments. Description of the Drawings

[0033] Figure 1 It is a schematic diagram of a dynamic allocation method for a 4G network provided by an embodiment of the present application. Detailed Embodiments

[0034] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0035] The terms "first", "second", etc. in the description and claims of this application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of this application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are usually of the same category, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the description and claims means at least one of the connected objects, and the character " / ", generally represents an "or" relationship between the associated objects before and after.

[0036] This embodiment provides a dynamic allocation method for a 4G network. When the terminal device is in a network connection state, multiple device network parameters are collected in real time; the device network parameters include multiple internal parameters, and the internal parameters of the device network parameters are input into a network model for inference and calculation to generate an inference and calculation result corresponding to each internal parameter; based on each inference and calculation result, dynamic programming is performed to obtain an optimal resource allocation strategy sequence; according to the optimal resource allocation strategy sequence, it is judged whether the current network connection state is in the optimal network connection state. If the judgment result is non-optimal, network switching is performed; after performing network switching, the storage method is planned, and the device network parameters, network model, inference and calculation results, dynamic programming, and optimal resource allocation strategy sequence are allocated to a reasonable storage space.

[0037] The following combines the accompanying drawings to detail the 4G network dynamic allocation method and medium provided by the embodiments of this application through specific embodiments and application scenarios.

[0038] As Figure 1 shown, this application provides a dynamic allocation method for a 4G network, and this method may include the following steps S100 to step S500.

[0039] Step S100, when the terminal device is in a network connection state, collects the network parameters of the device in real time.

[0040] Among them, the terminal device continuously monitors and collects key network parameters in real time through multiple configured monitoring modules to ensure the accuracy and reliability of subsequent network selection. The device network parameters include signal strength, network latency, and network congestion status. These device network parameters are monitored by the following modules respectively:

[0041] Optionally, a geographic location information monitoring module, which can obtain the geographic location information of the device in real time. Through the integrated GPS module or using network positioning services, the geographic location information monitoring module can provide accurate longitude and latitude data and output detailed geographic information of the current location as needed, including cities, streets, and location beacons, etc.

[0042] Optionally, there is a device network status information monitoring module which can monitor the current network connection status of the device in real time. This module can detect and output the current connected network type (such as LTE, Wi-Fi), signal quality, network switching history, and other relevant network status data so that users can understand the network connection situation of the device.

[0043] Optionally, the device network status information monitoring module includes a network latency monitoring module which can obtain network latency data in real time. By sending test data packets and measuring their round-trip time (RTT), the network latency monitoring module can accurately calculate the latency of the current network.

[0044] Optionally, there is a network congestion status monitoring module which evaluates the current network congestion situation in real time by analyzing the ratio of the actual data transmission rate to the theoretical maximum rate. This module can generate percentage data of network usage rate.

[0045] Optionally, there is a user behavior information monitoring module which can collect the behavior data of users when using the device. This module can record information such as the user's operation mode, frequently used applications, data usage, network access habits, etc., and provide detailed user behavior analysis to further optimize the allocation of network resources and user experience.

[0046] In this embodiment, through the collaborative work of the above modules, the terminal device can obtain and update key device network parameter data in real time. These data provide accurate and timely basic information for the subsequent construction of the static Bayesian network and dynamic programming, so as to ensure that the terminal device can always maintain the best network connection status in different network environments.

[0047] Step S200 inputs the device network parameters into the network model for inference calculation to generate an inference calculation result corresponding to each internal parameter.

[0048] Specifically, the preprocessing process can be decomposed into multiple steps. S201 establishes a Bayesian network for each of the device network parameters, where each node represents an internal parameter of a device network parameter, the edge represents the dependency relationship between nodes, and the nodes include parent nodes and child nodes.

[0049] Exemplarily, in this Bayesian network, the device network status information includes network latency, signal strength, and network congestion information. There is a dependency relationship between network latency and signal strength and network congestion. Specifically, network congestion may affect network latency, and signal strength may be affected by network congestion. Therefore, the node dependency relationship in the Bayesian network can be established as follows: Signal strength and network latency are child nodes of network congestion. Network congestion is the parent node of signal strength and network latency.

[0050] Furthermore, in S202, state definitions are created for the internal parameters in each node. The signal strength state levels include "strong, medium, weak", the network latency state levels include: "low latency, medium latency, high latency", and the network congestion state levels include: "unblocked, moderate, congested". A conditional probability table is constructed based on the states of each node.

[0051] Optionally, in this embodiment, step S203 is independent hypothesis verification, and the independent hypothesis verification uses a chi-square test to verify whether the conditional probability table is accurate.

[0052] Specifically, first, the sample data in the actual network operation collected through step S500 are obtained. These data include the observed values of signal strength, network latency, and network congestion at historical moments. The collected sample data are organized into an observed frequency table. According to the conditional probability table and the total number of samples, the expected frequency is calculated. Expected frequency = total number of samples × conditional probability.

[0053] The calculation formula for the chi-square statistic is:

[0054]

[0055] Where Oi is the observed frequency and Ei is the expected frequency. Substitute the observed frequency and expected frequency table item by item, calculate the chi-square value for each combination, and accumulate to obtain the total chi-square statistic.

[0056] Furthermore, the calculated chi-square statistic is compared with the critical value in the chi-square distribution table. According to the degrees of freedom and the significance level (usually 0.05), the corresponding critical value is found. If the chi-square statistic is greater than the critical value, the independence hypothesis is rejected, indicating that the conditional probability table needs to be adjusted because the observed data does not conform to the independence hypothesis. If the chi-square statistic is less than or equal to the critical value, the independence hypothesis is accepted, indicating that the conditional probability table is accurate and the Bayesian network model can continue to be used.

[0057] Exemplarily, in step S204, the joint probability distribution of the nodes is calculated. The calculation formula for the joint probability distribution is:

[0058]

[0059] where P(X) is the joint probability distribution of the node, and X i represents the i-th node, i is the index of the node, n is the total number of nodes, and Pa(X i ) is the parent node of node X i , and P(X i |Pa(X i )) is the conditional probability expression of the child node.

[0060] In this embodiment, it can be expressed as

[0061] P(Signal Strength, Network Delay, Degree of Network Congestion) = P(Degree of Network Congestion) × P(Signal Strength

[0062] |Degree of Network Congestion) × P(Network Delay|Degree of Network Congestion)

[0063] Furthermore, in step S205, the posterior probability is calculated according to the joint probability distribution. The calculation of the posterior probability includes applying marginalization to the joint probability distribution to convert the joint probability distribution into the posterior probability, and the posterior probability is the result of the inference calculation.

[0064] Exemplarily, the probability after marginalizing the signal strength can be expressed as:

[0065]

[0066] where: P(Network Delay = Low Delay) is the probability that the network delay is low delay, represents the sum over all possible degrees of network congestion, and P(Degree of Network Congestion = c, Network Delay = Low Delay) represents the joint probability when various degrees of network congestion and the network delay is low delay.

[0067] Exemplarily, P(Degree of Network Congestion = Moderate|Network Delay = Low Delay) is the posterior probability that the degree of network congestion is moderate when the network delay is low delay, and the posterior probability is the result of the inference calculation.

[0068] It can be understood that for geographical location information and user behavior information, the corresponding inference calculation results can be obtained through the above S201 to S205, which will not be elaborated here.

[0069] Step S300 processes the inference calculation result using dynamic programming. The processing process includes S301 state definition, S302 state transition, and S303 obtaining the optimal resource allocation strategy sequence. The specific dynamic programming can be the following steps:

[0070] In this embodiment, S301 is state definition, and the state definition is divided into state S(t), decision a(t), and reward R(t, a(t)). State S(t) represents the network resource allocation at time t. The decision represents the resource allocation strategy at time t, and the resource allocation strategy is based on the result of inference calculation. Reward R(t, a(t)) represents the immediate reward brought by selecting decision a(t).

[0071] Exemplarily, state S(t) can represent the bandwidth allocation at time t; decision a(t) can represent how much bandwidth to allocate to a certain type of user at time t; reward R(t, a(t)) can represent the negative network congestion level or the positive user satisfaction level.

[0072] Furthermore, S302 is state transition. State transition transfers the current state to the next state through the state transition equation. The state transition equation is:

[0073] S(t + 1) = f(S(t), a(t))

[0074] where S(t + 1) is the next state and f is the state transition function.

[0075] Among them, the state transition function can use linear functions, non - linear functions, etc. In this embodiment, since the inference calculation result obtained from the Bayesian network is processed, it is more appropriate to select a probability distribution function.

[0076] Furthermore, S303 is to obtain the optimal resource allocation strategy sequence. The Bellman equation is used to obtain the optimal resource allocation strategy sequence. The Bellman equation is:

[0077]

[0078] where V(t, S(t)) is the maximum cumulative reward at time t, γ is the discount factor, is the optimal decision;

[0079] Through recursive calculation, the optimal resource allocation strategy sequence is obtained according to the maximum cumulative reward.

[0080] Specifically, the recursive calculation starts from the initial state S(0) and the number of time steps T. Recursively calculate from t = T - 1, and recursively calculate the value of the Bellman equation V(t, S(t)) until t = 0. By maximizing the Bellman equation, find the optimal strategy a * (t) at each time step t, and finally obtain the optimal resource allocation strategy sequence {a * (0), a * (1), …, a * (T - 1)}.

[0081] In step S400, it is determined whether the current network connection state is in the optimal network connection state through the optimal resource allocation policy sequence. If it is not in the optimal network connection state, network switching is performed.

[0082] Specifically, the optimal network connection at each moment will be determined by the optimal resource allocation policy sequence.

[0083] In step S500, after performing network switching, the storage method is planned, and the device network parameters, network model, inference calculation results, dynamic programming, and optimal resource allocation policy sequence are allocated to reasonable storage spaces.

[0084] Exemplarily, different storage units will be divided from different storage media in the device. NOR Flash has a relatively fast read speed and good random read performance, and is suitable for storing small data blocks or codes that need to be frequently accessed. It can be allocated as a fast read unit and a random read unit in the system; NAND Flash has a high storage density and low cost, but poor random read performance, and is suitable for sequential read and write operations. It can be allocated as a sequential read unit in the system; eMMC is a NAND Flash storage device integrated with a controller, which supports both sequential and random reads, and has a large storage space, and different storage units can be allocated.

[0085] It should be noted that in this article, the term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method, article or device including that element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in a reverse order according to the functions involved. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0086] Through the description of the above embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM) and includes several instructions for causing a terminal (which can be a mobile phone, a computer, a server or a network device, etc.) to execute the methods of various embodiments of the present application.

[0087] The embodiments of the present application have been described above in conjunction with the accompanying drawings. However, the present application is not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative and not restrictive. Under the inspiration of the present application, those of ordinary skill in the art can also make many forms without departing from the purpose of the present application and the scope protected by the claims, and all of them belong to the protection scope of the present application.

Claims

1. A dynamic allocation method for a 4G network, characterized in that: The method comprises: S100 collects multiple device network parameters in real time when the terminal device is in a network connection state, wherein the device network parameters include multiple internal parameters; S200: Input the internal parameters of the device network parameters into the network model for inference calculation, and generate inference calculation results corresponding to each of the internal parameters; S300 performs dynamic programming based on each of the inference calculation results to obtain an optimal resource allocation strategy sequence; S400: judging whether the current network connection state is in an optimal network connection state according to the optimal resource allocation strategy sequence, and if the judgment result is not optimal, performing network switching; After executing the network switching, S500 plans a storage method and allocates the device network parameters, network model, reasoning calculation results, dynamic planning, and optimal resource allocation strategy sequence to a preset storage space.

2. The method for dynamic allocation of 4G network according to claim 1, characterized in that In S100, the internal parameters include geographic location information, device network status information and user behavior information, and the geographic location information, device network status information and user behavior information are all collected or read by the terminal device.

3. The dynamic allocation method of 4G network according to claim 2, characterized in that: The device network status information includes network delay, and the method for obtaining the network delay is: obtaining the network delay data by sending a test data packet and measuring its round-trip time.

4. The method for dynamic allocation of 4G network according to claim 2, characterized in that: The device network status information includes a network congestion status. The method for obtaining the network congestion status is: obtaining the network congestion status value by analyzing the ratio of an actual data transmission rate to a theoretical maximum rate.

5. The method for dynamic allocation of 4G network according to claim 1, characterized in that In S200, the network model is a static Bayesian network, and the reasoning calculation process includes: S201 establishes a Bayesian network for each of the device network parameters, where each node represents an internal parameter of the device network parameter, and an edge represents a dependency relationship between nodes, and the node includes a parent node and a child node; S202 constructs a conditional probability table, wherein the conditional probability table includes a priori probabilities or conditional probabilities of each of the nodes under different states, wherein the priori probabilities are represented in the conditional probability table as node probability distributions given a parent node, and the conditional probabilities are represented in the conditional probability ratio as probability distributions of the child nodes under different states; S203 verifies whether the conditional probability table is accurate by independent hypothesis verification, wherein the independent hypothesis verification uses a chi-square test; S204 calculates the joint probability distribution of the nodes, and the calculation formula of the joint probability distribution is: Where P(X) is the joint probability distribution of the nodes, X i represents the i-th node, i is the node index, n is the total number of nodes, Pa(X i ) is node X i The parent node of P(X i |Pa(X i )) is the child node conditional probability expression; S205 calculates the posterior probability according to the joint probability distribution, wherein the calculation of the posterior probability includes applying a marginalization method to the joint probability distribution to remove unnecessary node conditional probabilities and converting them into posterior probabilities, and using the posterior probability as the inference calculation result.

6. The method for dynamic allocation of 4G network according to claim 1, characterized in that In S300, the dynamic planning includes: The network resource allocation state, decision and benefit are defined, wherein the decision is based on the inference calculation result; the state transfer is performed through the state transfer equation; the maximum cumulative benefit is calculated recursively using the Bellman equation, thereby obtaining the optimal resource allocation strategy sequence.

7. The method for dynamic allocation of 4G network according to claim 1, characterized in that In S500, the storage space includes: a random read unit, a sequential read unit and a fast read unit. According to the data volume and the reading frequency, the device network parameters, network model, reasoning calculation results, dynamic programming and optimal resource allocation strategy sequence are allocated to the random storage unit, the sequential read unit and the fast read unit.

8. A readable storage medium, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by the processor, the steps of the dynamic allocation method of the 4G network as described in any one of claims 1 to 7 are implemented.

9. A readable storage medium according to claim 8, characterized in that: The storage medium includes NAND Flash, and the NAND Flash is allocated as the sequential reading unit in the system.

10. The readable storage medium according to claim 8, characterized in that: The storage medium includes NOR Flash, and the NOR Flash is allocated as the fast read unit and the random read unit in the system.

Images