A heterogeneous network dual-mode adaptive handover method and system

Abstract

The present application relates to the technical field of network switching, in particular to a heterogeneous network dual-mode adaptive switching method and system. The present application realizes dynamic response of network state to real-time change by acquiring multi-dimensional environment perception data and initial signal strength and calibrating signal strength in dynamic environment. By combining the calibrated signal strength, the connection state characteristics and side network consumption indicators of the dual-mode network are obtained, and the switching demand can be evaluated based on these data. This multi-dimensional evaluation mechanism can more accurately judge the switching demand, avoiding the limitations of traditional methods which only rely on signal strength. By generating an optimized directed time sequence graph and a network performance data set, the device can make scientific and reasonable switching decisions based on real-time network state snapshot data and predetermined optimized time sequence graph, dynamically adapt to the connection demand of the device under changing network conditions, thereby maintaining the best network experience and improving the network connection quality.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of network switching, in particular to a heterogeneous network dual-mode adaptive switching method and system. BACKGROUND

[0002] In the field of communication technology, heterogeneous network dual-mode refers to a device supporting two different types, different architectures or different technical standards of network modes (such as 4G and Wi-Fi), and can realize data transmission, connection management and adaptive switching between the two network modes. The core is to solve the limitations of single network in coverage range, transmission rate, power consumption control, scene adaptation, etc. through dual-mode compatibility and collaborative work, which is commonly used in mobile terminals, Internet of Things devices, industrial communication and other scenarios. The heterogeneous network dual-mode adaptive switching method refers to a method of automatically adjusting the access mode between multiple network modes according to the changes of network conditions (such as signal strength, bandwidth, delay, load, etc.). It can adaptively select the best network access mode according to different network conditions.

[0003] At present, the existing technology for the switching mode of heterogeneous network dual-mode mainly judges whether it needs to switch to another network by monitoring whether the current signal strength of the device network reaches the corresponding preset signal strength threshold in real time. However, the above-mentioned method is difficult to realize adaptive switching decision according to the dynamically changing environment and device state, and cannot match the real-time network and device demand, thereby affecting the network connection quality. SUMMARY

[0004] The main purpose of the present application is to provide a heterogeneous network dual-mode adaptive switching method to solve the technical problems in the prior art.

[0005] The present application provides a heterogeneous network dual-mode adaptive switching method, which comprises:

[0006] Obtain multi-dimensional environment perception data and initial signal strength of the dual-mode network in the heterogeneous network dual-mode coverage area, and dynamically calibrate the initial signal strength according to the multi-dimensional environment perception data to obtain calibrated signal strength;

[0007] Obtain the connection state characteristics and side network consumption indicators of the dual-mode network under the calibrated signal strength, and obtain the switching demand evaluation value according to the connection state characteristics and side network consumption indicators;

[0008] Determine the switching demand level of the dual-mode network according to the switching demand evaluation value, and obtain a plurality of signal fluctuation parameters and device power consumption change parameters corresponding to the switching demand level;

[0009] Generate time decision sequences of the dual-mode network according to the multiple signal fluctuation parameters and the device power consumption change parameters, and obtain network state snapshot data of the dual-mode network at multiple continuous time points in each time decision sequence;

[0010] Obtain corresponding directed time sequence graphs according to the multiple network state snapshot data, and obtain corresponding optimized time sequence graphs according to each directed time sequence graph;

[0011] Obtain a corresponding network performance data set according to each time decision sequence, and generate a dual-mode adaptive switching model according to the multiple network performance data sets and the optimized time sequence graphs, so that the device performs dual-mode network switching according to the dual-mode adaptive switching model.

[0012] As a preferred, the step of dynamically calibrating the initial signal strength according to the multi-dimensional environment perception data to obtain the calibrated signal strength comprises:

[0013] Obtain static environment features and dynamic interference factors in the multi-dimensional environment perception data, wherein the static environment features include geographic topology data and network infrastructure layout data, and the dynamic interference factors include fixed interference source data and dynamic interference source data;

[0014] Obtain building distribution data, wall material data and terrain elevation data according to the geographic topology data, and obtain shielding attenuation coefficients and terrain attenuation coefficients in the signal propagation process according to the building distribution data, wall material data and terrain elevation data;

[0015] Obtain the basic path loss parameter in the signal propagation process according to the network infrastructure layout data, and obtain the propagation loss coefficient of the dual-mode network according to the basic path loss parameter, the shielding attenuation coefficient and the terrain attenuation coefficient;

[0016] Obtain the interference loss coefficient according to the fixed interference source data and the dynamic interference source data, and obtain the total loss influence factor according to the interference loss coefficient and the propagation loss coefficient;

[0017] Calibrate the initial signal strength according to the total loss influence factor to obtain the calibrated signal strength.

[0018] As a preferred, the step of obtaining the interference loss coefficient according to the fixed interference source data and the dynamic interference source data comprises:

[0019] Obtain the fixed signal strength, the fixed center frequency and the fixed interference frequency bandwidth of each fixed interference source according to the fixed interference source data, and obtain the corresponding fixed effective interference frequency bandwidth according to each fixed center frequency and fixed interference frequency bandwidth;

[0020] According to each of the fixed effective interference frequency bandwidth, the fixed interference frequency bandwidth and the fixed signal strength, a corresponding fixed effective interference power is obtained, and according to a plurality of the fixed effective interference powers, a fixed total interference power is obtained;

[0021] According to the dynamic interference source data, a dynamic signal strength, a dynamic center frequency and a dynamic interference frequency bandwidth of each dynamic interference source are obtained, and according to each of the dynamic center frequency and the dynamic interference frequency bandwidth, a corresponding dynamic effective interference frequency bandwidth is obtained;

[0022] According to each of the dynamic effective interference frequency bandwidth, the dynamic interference frequency bandwidth and the dynamic signal strength, a corresponding dynamic effective interference power is obtained, and according to a plurality of the dynamic effective interference powers, a dynamic total interference power is obtained;

[0023] According to the fixed total interference power and the dynamic total interference power, a comprehensive interference power of the dual-mode network is obtained;

[0024] A useful signal strength of the dual-mode network is obtained, and according to the comprehensive interference power and the useful signal strength, an interference loss coefficient is obtained.

[0025] Preferably, the step of obtaining the switching demand evaluation value according to the connection state feature and the side network consumption index comprises:

[0026] A connection state feature is obtained, wherein the connection state feature comprises a connection stability sub-feature, a data transmission rate sub-feature, a connection duration sub-feature and a switching history success rate sub-feature;

[0027] According to the connection stability sub-feature, a connection stability comprehensive value is obtained, and according to the data transmission rate sub-feature, a data transmission rate comprehensive value is obtained;

[0028] According to the connection duration sub-feature, a connection duration comprehensive value is obtained, and according to the switching history success rate sub-feature, a switching history success rate comprehensive value is obtained;

[0029] According to the connection stability comprehensive value, the data transmission rate comprehensive value, the connection duration comprehensive value and the switching history success rate comprehensive value, a weighted sum is performed to obtain a connection state comprehensive value;

[0030] A side network consumption index is obtained, wherein the side network consumption index comprises a unit data transmission power consumption sub-feature, a bandwidth utilization rate sub-feature and a battery remaining power sub-feature;

[0031] According to the unit data transmission power consumption sub-feature, a unit power consumption comprehensive value is obtained, and according to the bandwidth utilization rate sub-feature, a bandwidth utilization rate comprehensive value is obtained;

[0032] According to the battery remaining power sub-feature, a battery power comprehensive value is acquired, and a consumption index comprehensive value is obtained by weighted summation of the unit power consumption comprehensive value, the bandwidth utilization rate comprehensive value and the battery power comprehensive value;

[0033] An application demand weight coefficient of the heterogeneous network is acquired, and a switching demand evaluation value of the heterogeneous network is acquired according to the application demand weight coefficient and the connection state comprehensive value and the consumption index comprehensive value.

[0034] Preferably, the step of generating a time decision sequence of the dual-mode network according to the plurality of signal fluctuation parameters and the device power consumption change parameters comprises:

[0035] The signal fluctuation parameters are acquired, wherein the signal fluctuation parameters comprise a signal strength change rate, a transmission delay fluctuation amplitude and a packet loss rate fluctuation amplitude;

[0036] The environment scene type is acquired, and the scene-based weight distribution is performed on each of the signal strength change rate, the transmission delay fluctuation amplitude and the packet loss rate fluctuation amplitude according to the environment scene type, to obtain a corresponding signal strength fluctuation weight, a delay fluctuation weight and a packet loss rate fluctuation weight;

[0037] The weighted summation is performed according to each of the signal strength change rate, the transmission delay fluctuation amplitude, the packet loss rate fluctuation amplitude, the signal strength fluctuation weight, the delay fluctuation weight and the packet loss rate fluctuation weight, to obtain a corresponding signal fluctuation comprehensive evaluation value;

[0038] The device power consumption change parameters are acquired, wherein the device power consumption change parameters comprise an instantaneous power consumption change rate, an average power consumption change rate and a peak power consumption change rate;

[0039] The scene-based weight distribution is performed on each of the instantaneous power consumption change rate, the average power consumption change rate and the peak power consumption change rate according to the environment scene type, to obtain a corresponding instantaneous power consumption weight, an average power consumption weight and a peak power consumption weight;

[0040] The weighted summation is performed according to each of the instantaneous power consumption change rate, the average power consumption change rate, the peak power consumption change rate, the instantaneous power consumption weight, the average power consumption weight and the peak power consumption weight, to obtain a corresponding power consumption change comprehensive evaluation value;

[0041] The corresponding network comprehensive scores are acquired according to each of the power consumption change comprehensive evaluation value and the signal fluctuation comprehensive evaluation value, and the plurality of network comprehensive scores are sorted in size order to obtain a time decision sequence.

[0042] Preferably, the step of acquiring a corresponding directed time sequence graph according to the plurality of network state snapshot data, and acquiring a corresponding optimized time sequence graph according to each of the directed time sequence graphs comprises:

[0043] According to each network state snapshot data, network performance characteristics and environment associated characteristics at the corresponding time are obtained, and according to each network performance characteristic and environment associated characteristic, a time characteristic vector at the corresponding time is constructed;

[0044] According to the time characteristic vectors of any two adjacent times, corresponding feature similarities are obtained, and whether each feature similarity is less than a preset similarity is judged;

[0045] If the feature similarity is not less than the preset similarity, it is determined that the network state of the two adjacent times corresponding to the feature similarity has no significant change, and a directed association edge between the two adjacent times needs to be established;

[0046] If the feature similarity is less than the preset similarity, it is determined that the network state of the two adjacent times corresponding to the feature similarity has a significant change, and a directed association edge between the two adjacent times does not need to be established;

[0047] Each time in each time decision sequence is defined as a graph node, and a plurality of graph nodes and directed association edges are organized in time sequence to form a corresponding directed time sequence graph;

[0048] The local density of each node in each directed time sequence graph is obtained, and whether each local density is less than a preset density is judged;

[0049] If the local density is less than the preset density, it is determined that the node is a redundant node, and each redundant node and its corresponding directed association edge are deleted from the directed time sequence graph to obtain a corresponding optimized time sequence graph.

[0050] The application also provides a heterogeneous network dual-mode adaptive switching system, comprising:

[0051] A dynamic calibration module is configured to obtain multi-dimensional environment perception data in a heterogeneous network dual-mode coverage area and initial signal strength of a dual-mode network, and to perform dynamic calibration on the initial signal strength according to the multi-dimensional environment perception data to obtain calibrated signal strength;

[0052] A first obtaining module is configured to obtain connection state characteristics and side network consumption indicators of the dual-mode network under the calibrated signal strength, and to obtain a switching demand evaluation value according to the connection state characteristics and the side network consumption indicators;

[0053] A determination module is configured to determine a switching demand level of the dual-mode network according to the switching demand evaluation value, and to obtain a plurality of signal fluctuation parameters and device power consumption change parameters corresponding to the switching demand level;

[0054] The first generation module is used for generating a time decision sequence of the dual-mode network according to the multiple signal fluctuation parameters and the device power consumption change parameter, and obtaining network state snapshot data of the dual-mode network at multiple continuous time instants in each time decision sequence;

[0055] The second acquisition module is used for acquiring a corresponding directed time sequence graph according to the multiple network state snapshot data, and acquiring a corresponding optimized time sequence graph according to each directed time sequence graph;

[0056] The second generation module is used for acquiring a corresponding network performance data set according to each time decision sequence, and generating a dual-mode adaptive switching model according to the multiple network performance data sets and the optimized time sequence graph, so that the device performs dual-mode network switching according to the dual-mode adaptive switching model.

[0057] Preferably, the dynamic calibration module comprises:

[0058] The first acquisition unit is used for acquiring static environment features and dynamic interference factors in the multi-dimensional environment perception data, wherein the static environment features include geographic topology data and network infrastructure layout data, and the dynamic interference factors include fixed interference source data and dynamic interference source data;

[0059] The second acquisition unit is used for acquiring building distribution data, wall material data and terrain elevation data according to the geographic topology data, and acquiring an occlusion attenuation coefficient and a terrain attenuation coefficient in a signal propagation process according to the building distribution data, the wall material data and the terrain elevation data;

[0060] The third acquisition unit is used for acquiring a basic path loss parameter in the signal propagation process according to the network infrastructure layout data, and acquiring a propagation loss coefficient of the heterogeneous network dual-mode according to the basic path loss parameter, the occlusion attenuation coefficient and the terrain attenuation coefficient;

[0061] The fourth acquisition unit is used for acquiring an interference loss coefficient according to the fixed interference source data and the dynamic interference source data, and acquiring a total loss influence factor according to the interference loss coefficient and the propagation loss coefficient;

[0062] The calibration unit is used for calibrating an initial signal strength according to the total loss influence factor to obtain a calibrated signal strength.

[0063] The application further provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor realizes the steps of the above-mentioned heterogeneous network dual-mode adaptive switching method when executing the computer program.

[0064] The application further provides a computer readable storage medium, which stores a computer program, and the computer program realizes the steps of the above-mentioned heterogeneous network dual-mode adaptive switching method when executed by a processor.

[0065] The application has the following beneficial effects: the application realizes dynamic response of network state to real-time changes by acquiring multi-dimensional environment perception data and initial signal strength and performing signal strength calibration in a dynamic environment, eliminates the dependence on fixed thresholds in traditional methods, can flexibly adjust according to actual environment and device state, thereby ensuring stable connection of the device under different network conditions, can acquire connection state features and side network consumption indicators of the dual-mode network by combining the calibrated signal strength, can evaluate switching requirements based on these data, this multi-dimensional evaluation mechanism can more accurately judge the switching requirements, avoids the limitation of traditional methods that only rely on signal strength, generates an optimized directed time sequence graph and a network performance data set, so that the device can make scientific and reasonable switching decisions according to real-time network state snapshot data and a predetermined optimized time sequence graph, can dynamically adapt to the connection requirements of the device under variable network conditions, thereby maintaining the best network experience and improving the network connection quality. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 The figure is a method flowchart of an embodiment of the application.

[0067] Figure 2 The figure is a system structure diagram of an embodiment of the application.

[0068] Figure 3 The figure is a computer device internal structure diagram of an embodiment of the application.

[0069] The implementation, functional features and advantages of the application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0070] It should be understood that the specific embodiments described herein are only used to explain the application and not to limit the application.

[0071] As shown in the figure, the application provides a heterogeneous network dual-mode adaptive switching method, which comprises: Figure 1

[0072] S1, multi-dimensional environment perception data in a heterogeneous network dual-mode coverage area and initial signal strength of the dual-mode network are acquired, and the initial signal strength is dynamically calibrated according to the multi-dimensional environment perception data to obtain calibrated signal strength;

[0073] S2, connection state features and side network consumption indicators of the dual-mode network under the calibrated signal strength are acquired, and a switching requirement evaluation value is acquired according to the connection state features and the side network consumption indicators.​

[0074] S3, determining a handover demand level of the dual-mode network according to the handover demand evaluation value, and obtaining a plurality of signal fluctuation parameters and device power consumption change parameters corresponding to the handover demand level;

[0075] S4, generating a time decision sequence of the dual-mode network according to the plurality of signal fluctuation parameters and device power consumption change parameters, and obtaining network state snapshot data of a plurality of continuous time instants in each time decision sequence of the dual-mode network;

[0076] S5, obtaining a corresponding directed time sequence graph according to a plurality of the network state snapshot data, and obtaining a corresponding optimized time sequence graph according to each of the directed time sequence graphs;

[0077] S6, obtaining a corresponding network performance data set according to each of the time decision sequences, and generating a dual-mode adaptive handover model according to a plurality of the network performance data sets and the optimized time sequence graph, so that the device performs dual-mode network handover according to the dual-mode adaptive handover model.

[0078] As described in the above steps S1-S2, wherein the core role of the multi-dimensional environment perception data is to describe the environmental basic conditions of the common coverage area of the dual-mode network, which has an impact on the signal propagation and interference of the two networks, etc., so it is not necessary to collect data separately according to the network, but to collect unified data for the entire coverage area. The initial signal strength of the dual-mode network includes the initial signal strength of the first network and the initial signal strength of the second network. The dynamic calibration of the initial signal strength by the multi-dimensional environment perception data is to independently calibrate the initial signal strength of the first network and the second network respectively through the multi-dimensional environment perception data. The purpose is to obtain the real and accurate real-time running state of the two types of networks. The parameters based on the independent calibration of the two types of networks are convenient for subsequent evaluation of switching demand and generation of time decision sequence. The switching demand evaluation value is based on the connection state characteristics and side network consumption index of the first network and the second network to obtain the first comprehensive performance evaluation value of the first network and the second comprehensive performance evaluation value of the second network. Based on the performance difference between the first comprehensive performance evaluation value and the second comprehensive performance evaluation value, it is judged whether it is greater than a preset threshold. For example, if the current access is the first network, if the second comprehensive performance evaluation value is greater than the first comprehensive performance evaluation value and the performance difference is greater than the preset threshold at this time, it means that the comprehensive performance of the second network is better, switching needs to be triggered, and the ratio of the first comprehensive performance evaluation value and the second comprehensive performance evaluation value is subtracted from 1 as the switching demand reference value. If the second comprehensive performance evaluation value is not greater than the first comprehensive performance evaluation value or the performance difference is not greater than the preset threshold, it means that the current comprehensive performance of the first network meets the demand and there is no need to switch, and the ratio of the second comprehensive performance evaluation value and the first comprehensive performance evaluation value is taken as the switching demand reference value. Finally, the switching demand evaluation value is calculated by the product of the switching demand reference value and the total correction coefficient.

[0079] The side network consumption index refers to a set of resource consumption parameters related to network connection and data transmission during the dual-mode access of the device to the heterogeneous network. The network state snapshot data refers to a set of data snapshots corresponding to the time-network state pairs collected at fixed time steps in the time decision sequence of the heterogeneous network dual mode. Traditional network switching methods usually rely on the current signal strength of the device to determine the switching requirement. However, this method does not take into account the dynamic changes in the environment, such as changes in signal strength due to weather, terrain, building obstructions, etc. This can cause the signal strength to reach a threshold at a certain moment, but it may not meet the actual needs of the device in practical applications. The present application can dynamically calibrate the initial signal strength by obtaining multi-dimensional environmental perception data, so that the signal strength can more accurately reflect the current connection quality of the device. Through dynamic calibration, the signal changes in complex environments can be more accurately reflected, avoiding incorrect switching decisions caused by sudden environmental changes. The network switching decision not only depends on the signal strength of the device, but also takes into account the real-time changes in the environment, making the switching more intelligent and adaptive to environmental changes. The prior art simply relies on signal strength to determine switching requirements, ignoring the connection state and resource consumption between the device and the network. Even if the signal strength reaches a threshold, the device may need more detailed switching judgment due to unstable network connection quality or device power consumption, etc. Therefore, the present application introduces connection state features and side network consumption indicators, which can not only evaluate switching requirements from the signal perspective, but also consider the overall performance indicators of the device and the network, making the switching decision more comprehensive and avoiding misjudgment when switching based solely on signal strength. By considering the connection state and side network consumption indicators, the switching timing can be adjusted flexibly according to the actual network load and device requirements, avoiding the problem of premature or late switching, thereby improving the connection stability of the network and the endurance of the device.

[0080] According to the above steps S3-S6, the step of obtaining a corresponding network performance data set according to each time decision sequence and generating a dual-mode adaptive switching model according to multiple network performance data sets and an optimized timing atlas includes: obtaining network running original data corresponding to each decision time in the time decision sequence, determining a target network of each time step according to the time decision sequence, and screening out running data of the target network from the network running original data, obtaining statistical values of network performance core indicators according to the screened target network running data, and obtaining a preliminary performance data set according to multiple statistical values, wherein the network performance core indicators include average transmission rate, average delay, average power consumption and switching success rate, removing and correcting abnormal data (such as instantaneous extreme value) in the preliminary performance data set to obtain a network performance data set, extracting performance feature vectors of each network performance data set and atlas structure features and decision correlation features of the optimized timing atlas, and constructing a model training sample set according to multiple performance feature vectors, atlas structure features and decision correlation features; dividing the model training sample set into a training set, a validation set and a test set, constructing a dual-mode adaptive switching initial model by using a graph neural network combined with a full connection layer, training the initial model using the training set, and adjusting model parameters through the validation set; inputting the test set into the trained model for performance verification, if the model prediction accuracy and decision matching degree meet a preset threshold, the final dual-mode adaptive switching model is determined; if not, additional training samples are added and retraining is performed until the model performance meets the standard, that is, the dual-mode adaptive switching model is obtained.

[0081] The step of determining the switching demand level of the dual-mode network according to the switching demand evaluation value includes obtaining a preset threshold interval, analyzing the matching relationship between the switching demand evaluation value and the preset threshold interval, and determining the switching demand level of the current heterogeneous network, wherein if the switching demand evaluation value is greater than the upper limit value of the preset threshold interval, the switching demand level of the current heterogeneous network is determined as high level, if the switching demand evaluation value is within the preset threshold interval, the switching demand level of the current heterogeneous network is determined as medium level, and if the switching demand evaluation value is less than the upper and lower limit values of the preset threshold interval, the switching demand level of the current heterogeneous network is determined as low level.

[0082] Traditional handover methods typically rely on simple signal strength assessments, neglecting the multi-layered nature of handover requirements. Different handover requirement levels correspond to different handover strategies, potentially requiring varying degrees of performance optimization or resource scheduling. This invention, however, determines handover requirement levels through handover requirement evaluation values. This ensures that handover depends not only on signal strength but also on a comprehensive consideration of connection quality and device status. Different handover requirement levels can trigger different levels of handover requirements. By using multiple signal fluctuation parameters and device power consumption variation parameters, appropriate measures are ensured based on different requirement levels. This hierarchical processing approach maximizes network performance and device resource utilization efficiency, avoiding... Unnecessary frequent switching or ignoring switching needs can be addressed by generating time-based decision sequences and acquiring network state snapshots from multiple consecutive moments at each time point. This enables more intelligent switching decisions. The network state snapshots provide a global perspective across multiple time periods, allowing network switching to dynamically respond to environmental changes rather than being limited to static decisions at a single moment. By capturing the changing trends of the network state, erroneous switching caused by decisions made at a single moment is avoided, improving the accuracy and stability of switching. By constructing a directed time series graph, the changing trends of the network state from one moment to another can be accurately captured, avoiding misjudgments caused by over-reliance on the local state at a single moment. The optimized timing graph is adjusted according to the network evolution pattern to optimize the handover timing. The optimized timing graph can provide a scientific basis for handover decisions, ensuring that the handover timing can maximize network connectivity and stability, and reduce latency and power consumption problems caused by frequent handover. The final handover model does not simply rely on a single signal strength or device power consumption, but integrates network performance data sets and optimized timing graphs. By comprehensively evaluating the network state, an adaptive handover model is generated, avoiding the limitations of a single data source and ensuring the accuracy and real-time performance of the model. The generated dual-mode adaptive handover model can flexibly adjust the handover strategy according to the real-time changes of devices and the network, and can realize personalized handover based on different application scenarios, environmental factors and device states. This allows devices to perform network handover at the optimal time, thereby ensuring the best connection quality and performance.

[0083] In one embodiment, step S1, which involves dynamically calibrating the initial signal strength based on the multi-dimensional environmental perception data to obtain the calibrated signal strength, includes:

[0084] S11. Obtain static environmental features and dynamic interference factors from the multi-dimensional environmental perception data, wherein the static environmental features include geographic topology data and network infrastructure layout data, and the dynamic interference factors include fixed interference source data and dynamic interference source data.

[0085] S12, obtaining building distribution data, wall material data and terrain elevation data according to the geographic topology data, and obtaining a shielding attenuation coefficient and a terrain attenuation coefficient in a signal propagation process according to the building distribution data, the wall material data and the terrain elevation data;

[0086] S13, obtaining a basic path loss parameter in a signal propagation process according to the network infrastructure layout data, and obtaining a propagation loss coefficient of a heterogeneous network dual mode according to a sum of the basic path loss parameter, the shielding attenuation coefficient and the terrain attenuation coefficient;

[0087] S14, obtaining an interference loss coefficient according to the fixed interference source data and the dynamic interference source data, and obtaining a total loss influence factor according to the interference loss coefficient and the propagation loss coefficient;

[0088] S15, calibrating an initial signal strength according to the total loss influence factor to obtain a calibrated signal strength.

[0089] As described in the above steps S11-S13, wherein the wall material data is used to obtain a signal penetration loss coefficient of the wall material in the frequency band of the dual mode network, the building distribution data is used to obtain the number of building walls on the signal propagation path, and the shielding attenuation coefficient is obtained by multiplying the number of building walls by the signal penetration loss coefficient, the terrain elevation data is used to obtain a terrain elevation difference and a terrain undulation frequency, and the terrain attenuation coefficient is obtained by adding the terrain elevation difference and the terrain undulation frequency after normalization and standardization, the network infrastructure layout data is used to obtain a horizontal distance between the device and the base station and a base station operating frequency, and the basic path loss parameter is calculated after normalization and standardization of the horizontal distance between the device and the base station and the base station operating frequency, wherein the calculation formula is: wherein B represents the basic path loss parameter, c represents the normalized horizontal distance between the device and the base station, and d represents the normalized base station operating frequency.

[0090] The prior art only relies on signal strength or certain aspect data for network switching judgment, and the present application can provide a comprehensive environmental background for subsequent calibration by obtaining static environmental features and dynamic interference factors of multi-dimensional environmental perception data, and by using multiple data sources in two dimensions of static and dynamic, not only can the environmental features be accurately understood, but also the factors that may interfere with the signal can be effectively identified, through this comprehensive analysis, the change of signal strength can be more comprehensively evaluated, the introduction of dynamic interference source data enables the entire system to adapt to real-time environmental changes, rather than just a static signal loss model, thereby providing more accurate dynamic calibration, by combining dynamic and static data, the problem of relying only on signal strength or other single data source is avoided, and the accuracy and adaptive ability of network switching are effectively improved, based on geographic topology data, building distribution, wall material and terrain elevation data are extracted, so as to facilitate the calculation of shielding attenuation and terrain attenuation of subsequent signal propagation, signal propagation is not only related to the physical environment, but also closely related to network infrastructure, by accurately obtaining network layout data, the present application can better reflect the actual influence of infrastructure on signal propagation, thereby improving the accuracy of signal strength calibration, by considering the layout of network infrastructure in the calculation of basic path loss, the network coverage range can be more accurately judged, thereby optimizing the switching decision, avoiding frequent switching due to insufficient signal strength, and ensuring the stability and quality of the network, real-time acquisition of shielding attenuation coefficient and terrain attenuation coefficient helps to adjust the signal strength in real time according to the current environment, so that the network switching decision is more in line with the actual demand, and the adaptive ability and switching accuracy of the network are improved.

[0091] As described in steps S14-S15, the interference loss coefficient is an important parameter for judging the influence of interference on the signal. By accurately obtaining the interference loss coefficient, the loss of the signal can be more accurately evaluated, and the false switching caused by interference can be avoided. The signal is affected by the interference source during transmission, which will cause additional loss. The existing technology mostly ignores the dynamic changing interference source, which easily leads to inaccurate prediction of signal strength. However, the present application can better reflect the influence of interference in the real environment by comprehensively considering fixed and dynamic interference sources, ensuring the accuracy of signal strength calibration. The dynamic interference source changes with time and location. The present application can more flexibly cope with changing environments by real-time collection of interference source data, improving the dynamic adaptability of the system. By comprehensively considering the interference loss and propagation loss, the total loss factor is obtained, thereby realizing more precise signal strength calibration. By comprehensively considering various loss factors, the present application can provide a more accurate signal loss evaluation, which helps to flexibly adjust the network switching threshold according to environmental changes, reduces unnecessary switching, and improves user experience. By considering environmental, interference, infrastructure and other factors, the present application can accurately dynamically calibrate the initial signal strength, avoiding the problem of large signal strength prediction error in the prior art. Precise signal strength calibration makes the network switching decision more scientific and reasonable, avoiding unnecessary switching caused by signal fluctuations or environmental changes, improving the stability and reliability of the network, and achieving more precise signal strength calibration, thereby greatly improving the accuracy and adaptability of network switching decision.

[0092] In one embodiment, the step S14 of obtaining the interference loss coefficient according to the fixed interference source data and the dynamic interference source data comprises:

[0093] S141, obtaining the fixed signal strength, the fixed center frequency and the fixed interference frequency bandwidth of each fixed interference source according to the fixed interference source data, and obtaining the corresponding fixed effective interference frequency bandwidth according to each fixed center frequency and fixed interference frequency bandwidth;

[0094] S142, calculating the corresponding fixed effective interference power according to each fixed effective interference frequency bandwidth, fixed interference frequency bandwidth and fixed signal strength, wherein the calculation formula is:

[0095]

[0096] Wherein, C represents the fixed effective interference power, h represents the fixed signal strength, D represents the fixed effective interference frequency bandwidth, and e represents the fixed interference frequency bandwidth.

[0097] S143, obtaining the fixed total interference power according to a plurality of fixed effective interference powers;

[0098] S144, acquiring a dynamic signal strength, a dynamic center frequency and a dynamic interference frequency bandwidth of each dynamic interference source according to the dynamic interference source data, and acquiring a corresponding dynamic effective interference frequency bandwidth according to each dynamic center frequency and dynamic interference frequency bandwidth;

[0099] S145, acquiring a corresponding dynamic effective interference power according to each dynamic effective interference frequency bandwidth, dynamic interference frequency bandwidth and dynamic signal strength, and acquiring a dynamic total interference power according to a plurality of dynamic effective interference powers;

[0100] S146, acquiring a comprehensive interference power of a dual-mode network according to the fixed total interference power and the dynamic total interference power;

[0101] S147, acquiring a useful signal strength of the dual-mode network, and calculating an interference loss coefficient according to the comprehensive interference power and the useful signal strength, wherein the calculation formula is:

[0102]

[0103] Wherein, E represents the interference loss coefficient, F represents the comprehensive interference power, and j represents the useful signal strength.

[0104] As described in steps S141-S143, wherein the fixed interference source refers to a device or apparatus that has a relatively fixed position and working state in the dual-mode network coverage area and continuously interferes with the network signal, such as a fixed wireless communication station, and the dynamic interference source refers to a device that has a position or working state that changes over time in the dual-mode network coverage area and dynamically interferes with the network signal, such as a mobile phone or computer, and the step of acquiring a corresponding fixed effective interference frequency bandwidth according to each fixed center frequency and fixed interference frequency bandwidth includes first determining whether the center frequency of each fixed interference source is within the working frequency range of the dual-mode heterogeneous network, if the center frequency is not within the working frequency range of the dual-mode heterogeneous network, the fixed interference source has no effective interference on the dual-mode heterogeneous network, and its fixed effective interference frequency bandwidth is 0, if the center frequency is within the working frequency range of the dual-mode heterogeneous network, the fixed interference source has effective interference on the dual-mode heterogeneous network, and the calculation formula of its fixed effective interference frequency bandwidth is:

[0105]

[0106] Wherein, D represents the fixed effective interference frequency bandwidth, e represents the fixed interference frequency bandwidth, g represents the fixed center frequency, and f min represents the minimum value of the working frequency of the dual-mode heterogeneous network, f maxThe working frequency maximum value of the heterogeneous network dual mode represents; the interference frequency bandwidth description signal occupies the frequency range width in the frequency domain, the unit is Hz (Hertz), and the fixed effective interference frequency bandwidth is the overlapping part of the interference source bandwidth and the heterogeneous network bandwidth.

[0107] The fixed total interference power is obtained by converting all the fixed effective interference powers of the heterogeneous network dual mode into linear power and then summing them up, and then converting back to logarithmic power. The calculation principles of dynamic total interference power, dynamic effective interference frequency bandwidth and dynamic effective interference power are the same as those of corresponding fixed total interference power, fixed effective interference frequency bandwidth and fixed effective interference power, respectively. By converting the fixed total interference power and the dynamic total interference power into linear power and then summing them up, the comprehensive interference power of the dual mode network can be obtained. The units of the fixed effective interference power and the fixed signal strength are dBm (decibel milliwatt). The fixed signal strength refers to the real-time signal strength of the fixed interference source. The signal strength here specifically refers to the received signal power strength, and the unit is dBm.

[0108] The prior art often only determines whether to switch according to the signal strength, ignoring the influence of the spectrum bandwidth on the interference. The present application can effectively reflect the actual influence of each fixed interference source on the network through the analysis of the signal strength, center frequency and frequency bandwidth of each fixed interference source. By obtaining the fixed effective interference frequency bandwidth, the effective use range of the interference spectrum can be more accurately determined, unnecessary signal processing is avoided, and the network's adaptability to interference and precise switching ability are improved. Compared with the prior art which relies on simple signal strength threshold determination, through more detailed spectrum analysis, the interference source can be better identified and evaluated, reducing misjudgment and delay. By calculating the fixed effective interference power, the actual interference degree of each fixed interference source on the network can be more intuitively quantified. By combining the two important factors of signal strength and frequency bandwidth, the calculation of interference power is more in line with the changes of the actual network environment, avoiding one-sided evaluation of the interference degree by simple signal strength. The present application can improve the accuracy of interference evaluation. The prior art often only focuses on a single interference source signal strength, ignoring the synergistic effect and overall interference influence of multiple interference sources. The present application integrates the effective interference powers of multiple fixed interference sources to provide the overall situation of interference in the fixed network. The calculation of total interference power helps to evaluate the overall interference level in the network and provides reliable data support for subsequent switching decisions.

[0109] As described in steps S144-S147 above, the prior art generally only focuses on the interference of a single network, ignoring the cross interference in a dual-mode network. The dynamic interference source often changes unpredictably in a heterogeneous network, so a more flexible and real-time interference analysis method is needed, which can more accurately capture the dynamically changing interference situation by monitoring the signal strength, center frequency and frequency bandwidth of the dynamic interference source in real time. The implementation of this step can accurately locate the interference in the dynamic network and differentiate it from static interference sources. Compared with the switching judgment in the prior art which only relies on fixed thresholds, the present application can dynamically adjust according to the actual changes in the network environment, improving the adaptive ability of the network. The calculation of the effective interference power of the dynamic interference source is the same as that of the static interference source, which considers multiple factors such as signal strength and frequency bandwidth. Dynamic interference sources usually exhibit strong time-varying behavior, so the calculation of their interference power needs to be able to track and calculate in real time. By obtaining the dynamic total interference power, the interference in the real-time environment can be fully quantified, further improving the accuracy of interference assessment. Compared with the simplified switching logic in the prior art, this method can provide more detailed and accurate interference power assessment. The calculation of the comprehensive interference power can combine fixed interference and dynamic interference to provide a more comprehensive interference analysis perspective for dual-mode network switching decisions. Through the calculation of the comprehensive interference power, it can help better predict the overall interference condition of the network and provide a more accurate basis for heterogeneous network switching decisions, avoiding signal quality fluctuations caused by network changes. By calculating the interference loss coefficient, the performance of the network in different interference environments can be fully evaluated, thereby providing more reliable parameters for switching decisions. By combining the comprehensive interference power and the useful signal strength to calculate the interference loss coefficient, not only can the network quality be more accurately measured, but also adaptive switching can be performed according to the actual situation to optimize the user experience and improve the communication quality and stability.

[0110] In one embodiment, the step S2 of obtaining a switching demand evaluation value according to the connection state feature and the side network consumption index comprises:

[0111] S21, obtain a connection state feature, wherein the connection state feature comprises a connection stability sub-feature, a data transmission rate sub-feature, a connection duration sub-feature and a switching history success rate sub-feature, and a connection stability comprehensive value is obtained by weighted summation according to the signal strength fluctuation frequency, signal strength continuous stable duration and connection interruption times of the connection stability sub-feature;

[0112] S22, obtaining a data transmission rate comprehensive value by weighted summation of the average transmission rate, the rate fluctuation coefficient and the rate valley value according to the data transmission rate sub-feature, and obtaining a connection duration comprehensive value by weighted summation of the current connection duration, the average connection duration under the same signal strength and the connection duration deviation rate according to the connection duration sub-feature;

[0113] S23, obtaining a handover history success rate comprehensive value by weighted summation of the same signal strength handover success rate and the same scene handover success rate according to the handover history success rate sub-feature, and obtaining a connection state comprehensive value by weighted summation of the connection stability comprehensive value, the data transmission rate comprehensive value, the connection duration comprehensive value and the handover history success rate comprehensive value;

[0114] S24, obtaining a side network consumption index, wherein the side network consumption index comprises a unit data transmission power consumption sub-feature, a bandwidth utilization rate sub-feature and a battery residual capacity sub-feature, and obtaining a unit power consumption comprehensive value by weighted summation of the average unit power consumption and the power consumption peak value according to the unit data transmission power consumption sub-feature;

[0115] S25, obtaining a bandwidth utilization rate comprehensive value by weighted summation of the average bandwidth utilization rate and the bandwidth utilization rate fluctuation amplitude according to the bandwidth utilization rate sub-feature, and obtaining a battery capacity comprehensive value by weighted summation of the current residual capacity, the capacity decline rate and the estimated residual use duration according to the battery residual capacity sub-feature;

[0116] S26, obtaining a consumption index comprehensive value by weighted summation of the unit power consumption comprehensive value, the bandwidth utilization rate comprehensive value and the battery capacity comprehensive value;

[0117] S27, obtaining an application demand weight coefficient of the heterogeneous network, and obtaining a comprehensive performance score by weighted summation of the connection state comprehensive value and the consumption index comprehensive value according to the application demand weight coefficient;

[0118] S28, obtaining historical deviation data of the connection stability comprehensive value, the data transmission rate comprehensive value, the connection duration comprehensive value, the handover history success rate comprehensive value, the unit power consumption comprehensive value, the bandwidth utilization rate comprehensive value and the battery capacity comprehensive value respectively, and obtaining a corresponding deviation correction coefficient according to each historical deviation data;

[0119] S29, obtaining a total correction coefficient by weighted summation of a plurality of the deviation correction coefficients, and obtaining a handover demand evaluation value of the heterogeneous network according to the total correction coefficient and the comprehensive performance score.

[0120] As described in steps S21-S23, in the calculation of the switching history success rate comprehensive value, the connection duration comprehensive value, the data transmission rate comprehensive value, and the connection stability comprehensive value, the respective corresponding calculation required parameters need to be standardized and normalized before being weighted and calculated. Similarly, in the calculation of the battery power comprehensive value, the bandwidth utilization rate comprehensive value, and the unit power consumption comprehensive value, the respective corresponding calculation required parameters need to be standardized and normalized before being weighted and calculated. The application demand weight coefficient is determined by the current application type of the device. The application type includes delay-sensitive type, power-sensitive type, and bandwidth-sensitive type. For example, when the current application type of the device is delay-sensitive type, the connection stability weight is 0.3, the data transmission rate weight is 0.3, the delay weight is 0.2, the power consumption weight is 0.1, and the bandwidth weight is 0.1.

[0121] The acquisition of the connection state characteristics analyzes the quality of the network connection from multiple dimensions, including connection stability, data transmission rate, connection duration, and switching history success rate. This comprehensive evaluation method can more accurately reflect the current connection quality and potential switching demand of the device than the traditional method of relying solely on signal strength. Specifically, the weighted sum of the connection stability can identify connections with frequent or unstable signal fluctuations, helping to determine whether there is a risk of switching. The data transmission rate and connection duration sub-features can provide real-time adjustment information for the device in a dynamically changing environment by monitoring rate fluctuations and connection duration, avoiding switching failures due to insufficient network bandwidth or connection duration. The switching history success rate reflects the past switching success rate of the device, allowing for an understanding of the device's switching ability under similar conditions through historical data, providing a reference for new switching decisions. Compared to the existing technology that relies solely on signal strength to determine switching demand, the weighted sum method combining multiple connection state characteristics can provide a comprehensive and dynamic evaluation system that more accurately reflects the current network quality and switching demand of the device, thereby reducing unnecessary switching or non-switching situations caused by misjudgment.

[0122] As described in steps S24-S26, the network consumption indicators reflect the energy efficiency performance of the device in the current connection state, covering power consumption, bandwidth utilization and battery remaining capacity. The comprehensive consideration of these indicators helps to balance network performance and energy efficiency requirements in the switching decision, specifically including the weighted sum of unit data transmission power consumption to effectively monitor the energy consumption of the device during data transmission, avoiding unnecessary network switching in high power consumption state, the analysis of bandwidth utilization to identify the utilization of current network resources, helping to judge whether the device has used the maximum bandwidth or whether it can improve bandwidth usage efficiency by switching to other networks, and the evaluation of battery remaining capacity to comprehensively consider the device power status, reducing the situation that the device is forced to switch at inappropriate time due to insufficient power. Compared with the prior art, the side network consumption indicators provide a new method to balance performance and energy efficiency. The previous switching method relies mainly on simple signal strength or bandwidth threshold, ignoring factors such as battery power and power consumption that affect the normal operation of the device. Through dynamic monitoring of these consumption indicators, the device can select the best switching time within a reasonable power consumption range, thereby prolonging the device's battery life and optimizing the user experience.

[0123] The traditional network switching method does not consider the difference in application requirements, and often simply takes network conditions as the dominant factor, ignoring the actual requirements of the application layer. The application requirement weight coefficient introduced in the present application adjusts the difference in application requirements in different network environments. By assigning appropriate weight coefficients to different network applications, the switching decision can be optimized according to the real-time requirements of the application (such as bandwidth, delay requirements, etc.), so that the system can determine whether to switch according to the specific requirements of the application, avoiding frequent switching in unnecessary situations, especially in the balance between high-bandwidth-demand applications and low-bandwidth-demand applications. By introducing the weight coefficient, the switching decision can be made more intelligently according to the application requirements, thereby optimizing the overall network performance. The present application obtains a comprehensive performance score by weighting and summing the connection state comprehensive value and the consumption index comprehensive value, which can comprehensively reflect the performance of the device in the current heterogeneous network environment, integrating the device connection quality and network consumption. The comprehensive score provides a basis for switching decisions. Compared with the traditional method, this calculation method is more flexible and comprehensive, avoiding biased judgments caused by only looking at one aspect (such as signal strength or bandwidth). By considering multiple indicators such as connection quality, network load, and device consumption, the network environment can be better adapted to dynamic changes, improving the accuracy of network switching decisions and avoiding performance degradation due to excessive focus on a single indicator. The present application introduces a historical deviation data correction coefficient, which can consider past errors and biases when making decisions, allowing the system to adapt to changes in the environment and differences in device characteristics. Different devices and environments may produce different network performance. By correcting historical deviation data, the system can dynamically adjust based on the past performance of the device, improving the accuracy of switching. Compared with the prior art, the correction mechanism based on historical deviation data makes the network switching decision more personalized, allowing for optimized adjustments based on the actual usage and historical performance of the device, reducing false switching caused by device differences, network environment changes, and other factors. The final switching requirement evaluation value is obtained by weighting and summing various feature data to give a global switching requirement evaluation result. This result is obtained by considering connection quality, consumption, application requirements, and historical deviation, and can intelligently determine whether to switch to another network based on the current state of the device and changes in the network environment. This evaluation method based on comprehensive features and historical data is more adaptable, avoiding misjudgments in dynamic network environments and ensuring stable device connection quality. It can provide a more accurate and intelligent switching decision scheme for devices in heterogeneous networks, fully considering dynamic environmental changes, device status, and application requirements in multiple dimensions, effectively improving the switching performance of heterogeneous networks, and maximizing network quality and device running efficiency.

[0124] In one embodiment, the step S4 of generating the time decision sequence of the dual-mode network according to the plurality of signal fluctuation parameters and the device power consumption change parameters comprises:

[0125] S41, acquiring signal fluctuation parameters, wherein the signal fluctuation parameters include signal strength change rate, transmission delay fluctuation amplitude and packet loss rate fluctuation amplitude;

[0126] S42, acquiring an environmental scene type, and respectively performing scene-based weight distribution on each of the signal strength change rate, the transmission delay fluctuation amplitude and the packet loss rate fluctuation amplitude according to the environmental scene type to obtain corresponding signal strength fluctuation weight, delay fluctuation weight and packet loss rate fluctuation weight;

[0127] S43, performing weighted summation according to each of the signal strength change rate, the transmission delay fluctuation amplitude, the packet loss rate fluctuation amplitude, the signal strength fluctuation weight, the delay fluctuation weight and the packet loss rate fluctuation weight to obtain a corresponding signal fluctuation comprehensive evaluation value;

[0128] S44, acquiring device power consumption change parameters, wherein the device power consumption change parameters include instantaneous power consumption change rate, average power consumption change rate and peak power consumption change rate;

[0129] S45, respectively performing scene-based weight distribution on each of the instantaneous power consumption change rate, the average power consumption change rate and the peak power consumption change rate according to the environmental scene type to obtain corresponding instantaneous power consumption weight, average power consumption weight and peak power consumption weight;

[0130] S46, performing weighted summation according to each of the instantaneous power consumption change rate, the average power consumption change rate, the peak power consumption change rate, the instantaneous power consumption weight, the average power consumption weight and the peak power consumption weight to obtain a corresponding power consumption change comprehensive evaluation value;

[0131] S47, performing weighted summation according to each of the power consumption change comprehensive evaluation value and the signal fluctuation comprehensive evaluation value to obtain a corresponding network comprehensive score, and sorting a plurality of the network comprehensive scores in order of size to obtain a time decision sequence.

[0132] As described in steps S41-S43, by judging whether each network comprehensive score is greater than the preset score, if the network comprehensive score is greater than the preset score, it is determined that the current heterogeneous network under the network score needs to be switched, otherwise it is not switched, the corresponding switching strategy is obtained, and then the switching strategy under each network comprehensive score is sorted to obtain the time decision sequence; the prior art often does not fully consider the difference of environmental factors, and a fixed threshold or a simple switching standard may be used, which may cause waste of performance in different environments; for example, a mobile device supports 5G (network A) and Wi-Fi6 (network B) dual-mode switching, and needs to generate a time decision sequence for the next 30 seconds (time step 1 second, a total of 30 decision points) through the above steps, the specific application process includes real-time network A and network B basic communication data through the built-in network monitoring module of the device, the sampling time interval is set to 1 second, 10 seconds of historical data have been continuously collected, which are used to calculate signal fluctuation parameters and power consumption change parameters, the current location is confirmed through the device positioning module, and the environment sensing data is matched from the preset scene library to match the current environment scene type, then the signal strength change rate, transmission delay fluctuation amplitude, packet loss rate fluctuation amplitude, signal strength fluctuation weight, delay fluctuation weight and packet loss rate fluctuation weight are obtained and weighted sum calculation is performed to obtain the signal fluctuation comprehensive evaluation value, and the instantaneous power consumption change rate, average power consumption change rate, peak power consumption change rate, instantaneous power consumption weight, average power consumption weight and peak power consumption weight are obtained and weighted sum calculation is performed to obtain the power consumption change comprehensive evaluation value, the signal fluctuation comprehensive evaluation value weight and the power consumption change comprehensive evaluation value weight are set, and calculation is performed to obtain the network comprehensive score, finally, for each time step (1 second), the comprehensive scores of the dual-mode networks are compared, and the network with higher score is selected as the switching decision at that time, assuming that the signal fluctuation and power consumption change parameters of network A and network B are stable (confirmed by LSTM prediction) in the next 30 seconds, after calculating the comprehensive score of each time step, the time decision sequence is obtained.

[0133] The traditional method usually only determines whether to switch the network according to a single signal strength threshold, which ignores the complex dynamic factors in the transmission process, such as delay fluctuation and packet loss phenomenon, and the present application can comprehensively analyze the stability and performance fluctuation of the current network by obtaining signal fluctuation parameters. The comprehensive consideration of these parameters is different from the traditional technology which only focuses on signal strength. It more comprehensively reflects the fluctuation of network quality. The signal strength change rate reflects the change trend of signal strength over time, which can predict the change of network quality in the short term and help to judge whether there is a short-term fluctuation risk. The transmission delay fluctuation amplitude directly affects the real-time performance of network response. The fluctuation of transmission delay can reflect the stability and processing capacity of the network. The packet loss rate fluctuation amplitude directly affects the data transmission quality. The fluctuation of packet loss rate can help to accurately capture the instantaneous change of network load and avoid performance degradation caused by packet loss. Considering the multi-dimensional parameters of signal strength, delay and packet loss can more comprehensively reflect the network performance and provide more accurate basis for subsequent network switching decision. The environment scene type (such as city, countryside, industrial area, etc.) will have important influence on signal propagation, delay and packet loss rate. According to different environment scenes, the signal fluctuation parameters are weighted. Not only the real-time data of network fluctuation is considered, but also the external factors of network operation are considered to interfere with the signal. Ensure that the weight distribution is more in line with the needs of the actual use scene. Different environment scenes will lead to different reasons for signal quality change. For example, in urban environment, signal strength may be subject to greater fluctuation due to high buildings or other equipment interference, while in rural areas, delay fluctuation may be more significant. Through scenario-based weight distribution, the evaluation standard can be dynamically adjusted to ensure more accurate switching decisions in different scenarios, so that the network switching mechanism can automatically adapt to the needs of different environments and avoid one-size-fits-all solutions. Through this weighted summation, multiple signal fluctuation factors can be integrated into a single evaluation value, which more accurately reflects the comprehensive performance of the current network state and can better provide scientific basis for subsequent decision-making, avoiding the one-sidedness of single signal strength judgment.

[0134] As described in steps S44-S47, the change of device power consumption is directly related to the energy efficiency and use time of the device, especially in battery-powered mobile devices. By considering the instantaneous change, average change and peak change of power consumption, more accurate power consumption evaluation can be achieved, providing more reference data for switching decisions. By introducing multiple power consumption change parameters, the energy consumption of the device can be comprehensively monitored, rather than relying solely on the change of instantaneous power consumption. This can better balance the endurance and network performance requirements of the device. The comprehensive evaluation of instantaneous power consumption, average power consumption and peak power consumption can avoid the influence of single parameter fluctuations on device operation efficiency, thereby improving the rationality of switching decisions and avoiding additional energy consumption caused by frequent switching. By weighting and summing multiple power consumption change parameters for comprehensive evaluation, the overall energy efficiency performance of the device can be more accurately reflected. By comprehensively evaluating signal fluctuations and power consumption changes, the network comprehensive score considers not only the network quality but also the power consumption efficiency of the device. By sorting multiple network scores, the optimal network can be automatically selected according to time and environmental changes, realizing adaptive switching. Unlike traditional technologies that use fixed thresholds to determine network switching, the present application comprehensively evaluates the signal and power consumption quality of the network to ensure that the switching decision considers both performance and energy efficiency. The sorting of multiple network comprehensive scores provides a dynamic adjustment mechanism that can achieve optimal switching decisions according to different environments and device states, thereby effectively improving network connection quality.

[0135] In one embodiment, the step S5 of obtaining a corresponding directed time sequence graph according to a plurality of network state snapshot data and obtaining a corresponding optimized time sequence graph according to each directed time sequence graph comprises:

[0136] S51, obtaining network performance characteristics and environmental correlation characteristics at each time according to each network state snapshot data, and constructing a time feature vector at each time according to each network performance characteristic and environmental correlation characteristic;

[0137] S52, calculating the feature similarity of the time feature vectors of any two adjacent times according to the cosine similarity, and determining whether each feature similarity is less than a preset similarity;

[0138] If the feature similarity is not less than the preset similarity, it is determined that the network state of the two adjacent times corresponding to the feature similarity has no significant change, and a directed correlation edge between the two adjacent times needs to be established.

[0139] If the feature similarity is less than the preset similarity, it is determined that the network state of the two adjacent times corresponding to the feature similarity has significant change, and a directed correlation edge between the two adjacent times does not need to be established.

[0140] S53, define each time point in each time decision sequence as a graph node, and organize a plurality of the graph nodes and directed association edges in time sequence to form a corresponding directed time sequence graph;

[0141] S54, obtain a local density of each node in each directed time sequence graph, and determine whether each local density is less than a preset density;

[0142] If the local density is less than the preset density, determine that the node is a redundant node, and delete each redundant node and its corresponding directed association edge from the directed time sequence graph to obtain a corresponding optimized time sequence graph.

[0143] As described in steps S51-S52, wherein the network performance features include signal stability features, rate fluctuation features, delay target features, low power consumption features, and packet loss rate level features, the environment association features include location association features and interference influence features, the local density is obtained by summing the feature similarities of the node and all nodes directly connected thereto, for a pair of nodes disconnected after deleting a redundant node (e.g., t and t+2, and the original intermediate node t+1 is a redundant node), the feature similarity is recalculated, and if the feature similarity is not less than a preset similarity, a directed association edge is supplemented to avoid graph fragmentation; the directed time sequence graph is a directed time sequence graph constructed based on network state snapshot data of a dual-mode network at continuous time points, through feature extraction, node definition, and edge weight calculation, and its core role is to intuitively reflect the change law, association relationship, and switching decision association information of the network state over time, to provide structured environmental feature support for subsequent optimized time sequence graph construction and dual-mode adaptive switching model training, in the entire heterogeneous network dual-mode adaptive switching method, the directed time sequence graph is a key link between network state data collection and optimized time sequence graph construction, and its role includes converting discrete, multi-dimensional network state snapshot data into a structured graph of nodes-edges, solving the problem of chaotic original data dimensions that are difficult to directly use for model training, intuitively displaying the association strength of network states at adjacent time points through edge weights, and in the subsequent optimized time sequence graph construction step, the directed time sequence graph will serve as basic data to finally generate an optimized graph, and provide high-quality input features for the dual-mode adaptive switching model.

[0144] The present application can ensure real-time and comprehensive analysis of network state by obtaining network performance features and environment correlation features at each moment, and by introducing multi-dimensional data, considering environmental changes, device states and other factors, which makes the network switching decision more intelligent and accurate. The environment correlation features can effectively make up for the shortcomings of signal strength as the only criterion. For example, signal strength may fluctuate due to geographical environment, while environmental features such as temperature, humidity, and device load can help the system better adapt to dynamic environmental changes. The feature vector can effectively represent the network state and environmental state at each moment, facilitating subsequent analysis and processing. This approach ensures the accuracy and consistency of the moment data and avoids instability caused by relying solely on signal strength. The use of cosine similarity can accurately quantify the feature similarity of any two adjacent moments. By calculating the angle difference between feature vectors, the significance of network state changes can be determined. This approach can adapt to different network environment changes. By using cosine similarity, false positives caused by simply using threshold values to determine whether changes have occurred can be avoided. For example, in some cases, even if the signal strength changes slightly, the network state may not have changed substantially. Cosine similarity can more accurately reflect this. Cosine similarity allows dynamic adjustment of the similarity threshold at different time periods or in different environments, making the network state change judgment more flexible and adaptable to the dynamic characteristics of the network environment. The present application introduces a method of determining whether to establish a directed correlation edge based on similarity, which solves the problem of switching decisions relying solely on signal strength thresholds in traditional methods. With this judgment mechanism, network switching not only considers signal strength but also takes into account the significance of network state changes, making the switching decision more intelligent. In the case of no significant change in network state between adjacent moments, no directed correlation edge is established, avoiding frequent and meaningless network switching, which helps to reduce unnecessary overhead and delay caused by network switching and ensures the stability of network connection. By judging similarity, the system can dynamically adapt to changes in the network environment rather than adhering to pre-set rules. This adaptive feature allows the system to optimize switching decisions in different environments and network conditions.

[0145] As described in steps S53-S54 above, the application represents the time sequence relationship by taking each time point as a graph node and using a graph structure, which helps to better understand the change rule of network state over time. The graph-based representation method makes the whole network state change process more visualized, facilitating subsequent analysis and decision-making. The graph nodes and directed edges are organized in chronological order, which can effectively capture the time dependence of network state changes and ensure that the network switching decision can consider the time dimension factor. By calculating the local density, redundant graph nodes can be identified, which represent time points that do not significantly contribute to network switching decisions. By deleting these redundant nodes, the graph structure is optimized, unnecessary calculations and decision-making processes are reduced, and the system efficiency is improved. Deleting redundant nodes can effectively simplify the graph structure, reduce the computational burden, and make subsequent analysis and decision-making more efficient. Compared to the traditional method of processing complex and lengthy time series data, the optimized time sequence graph is more helpful for quickly making network switching decisions. By eliminating redundant nodes, this method can more efficiently utilize computing resources and network resources, avoiding unnecessary data processing and analysis, and improving the overall system efficiency and response speed. The final optimization step optimizes the graph structure by deleting redundant nodes and edges, simplifies the subsequent analysis process, and effectively removes parts that have no substantial impact on network switching decisions, improving system efficiency and response speed. By deleting redundant nodes and edges, the complexity of the graph is reduced, enabling the system to process data and make switching decisions more quickly. It can handle large-scale time series data without sacrificing system performance and accuracy. The optimized graph structure can more accurately reflect the time points that have a practical impact on network switching, avoiding the computational burden caused by overly complex models, and enabling the system to better adapt to changes in dynamic environments.

[0146] As Figure 2 shown, the application also provides a heterogeneous network dual-mode adaptive switching system, comprising:

[0147] A dynamic calibration module is configured to obtain multi-dimensional environment perception data and initial signal strength of a dual-mode network within a dual-mode coverage area of a heterogeneous network, and dynamically calibrate the initial signal strength based on the multi-dimensional environment perception data to obtain calibrated signal strength.

[0148] A first obtaining module is configured to obtain connection state features and side network consumption indicators of the dual-mode network under the calibrated signal strength, and obtain a switching demand evaluation value based on the connection state features and the side network consumption indicators.

[0149] A determination module is configured to determine a switching demand level of the dual-mode network based on the switching demand evaluation value, and obtain a plurality of signal fluctuation parameters and device power consumption change parameters corresponding to the switching demand level.

[0150] The first generating module is configured to generate a time decision sequence of the dual-mode network according to the plurality of signal fluctuation parameters and the device power consumption change parameter, and obtain network state snapshot data of the dual-mode network at a plurality of continuous time instants in each time decision sequence;

[0151] The second obtaining module is configured to obtain a corresponding directed time sequence graph according to the plurality of network state snapshot data, and obtain a corresponding optimized time sequence graph according to each directed time sequence graph;

[0152] The second generating module is configured to obtain a corresponding network performance data set according to each time decision sequence, and generate a dual-mode adaptive switching model according to the plurality of network performance data sets and the optimized time sequence graph, so that the device performs dual-mode network switching according to the dual-mode adaptive switching model.

[0153] In one embodiment, the dynamic calibration module comprises:

[0154] The first obtaining unit is configured to obtain static environment features and dynamic interference factors in the multi-dimensional environment perception data, wherein the static environment features include geographic topology data and network infrastructure layout data, and the dynamic interference factors include fixed interference source data and dynamic interference source data;

[0155] The second obtaining unit is configured to obtain building distribution data, wall material data, and terrain elevation data according to the geographic topology data, and obtain an occlusion attenuation coefficient and a terrain attenuation coefficient in a signal propagation process according to the building distribution data, the wall material data, and the terrain elevation data;

[0156] The third obtaining unit is configured to obtain a basic path loss parameter in a signal propagation process according to the network infrastructure layout data, and obtain a propagation loss coefficient of the heterogeneous network dual-mode according to the basic path loss parameter, the occlusion attenuation coefficient, and the terrain attenuation coefficient;

[0157] The fourth obtaining unit is configured to obtain an interference loss coefficient according to the fixed interference source data and the dynamic interference source data, and obtain a total loss impact factor according to the interference loss coefficient and the propagation loss coefficient;

[0158] The calibration unit is configured to calibrate the initial signal strength according to the total loss impact factor to obtain a calibrated signal strength.

[0159] It should be noted that each module and unit in the heterogeneous network dual-mode adaptive switching system corresponds to each step in the heterogeneous network dual-mode adaptive switching method.

[0160] As shown in Figure 3 The computer device can be a server, and its internal structure can be as shown inFigure 3 The computer device shown in the figure includes a processor, a memory, a network interface and a database connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The database of the computer device is configured to store all data required by the process of the heterogeneous network dual-mode adaptive handover method. The network interface of the computer device is configured to communicate with an external terminal through a network connection. The computer program is executed by the processor to implement the heterogeneous network dual-mode adaptive handover method.

[0161] Those skilled in the art can understand that, Figure 3 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied.

[0162] The embodiment of the present application also provides a computer readable storage medium, which stores a computer program. The computer program is executed by the processor to implement any one of the above-mentioned heterogeneous network dual-mode adaptive handover methods.

[0163] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer readable storage medium. When the computer program is executed, it can include the processes of the above-mentioned embodiment methods. Any reference to the memory, storage, database or other medium provided by the present application and used in the embodiments can include non-volatile and / or volatile memory. The non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. The volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM) and the like.

[0164] It is to be understood that the terminology "including", "comprising", or any other variation thereof, is intended to cover a non-exclusive inclusion such that process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0165] The above description is merely the preferred embodiments of the present application, and is not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made according to the content of the present application specification and drawings, or directly or indirectly applied to other related technical fields, is also included in the patent protection scope of the present application.

Claims

1. A method for dual-mode adaptive handover in a heterogeneous network, the method comprising: The method comprises the following steps: acquiring multi-dimensional environment perception data and initial signal strength of a dual-mode network in a dual-mode coverage area of a heterogeneous network, and dynamically calibrating the initial signal strength according to the multi-dimensional environment perception data to obtain calibrated signal strength; acquiring connection state characteristics and side network consumption indicators of the dual-mode network under the calibrated signal strength, and acquiring a handover demand evaluation value according to the connection state characteristics and the side network consumption indicators; determining a handover demand level of the dual-mode network according to the handover demand evaluation value, and acquiring a plurality of signal fluctuation parameters and device power consumption change parameters corresponding to the handover demand level; generating a time decision sequence of the dual-mode network according to the plurality of signal fluctuation parameters and device power consumption change parameters, and acquiring network state snapshot data of the dual-mode network at a plurality of continuous time instants in each time decision sequence; acquiring a corresponding directed time sequence graph according to a plurality of the network state snapshot data, and acquiring a corresponding optimized time sequence graph according to each of the directed time sequence graphs; acquiring a corresponding network performance data set according to each of the time decision sequences, and generating a dual-mode adaptive handover model according to a plurality of the network performance data sets and the optimized time sequence graph, so that the device performs dual-mode network handover according to the dual-mode adaptive handover model.

2. The heterogeneous network dual-mode adaptive handover method of claim 1, wherein, The step of dynamically calibrating the initial signal strength according to the multi-dimensional environment perception data to obtain the calibrated signal strength comprises: acquiring static environment characteristics and dynamic interference factors in the multi-dimensional environment perception data, wherein the static environment characteristics include geographic topology data and network infrastructure layout data, and the dynamic interference factors include fixed interference source data and dynamic interference source data; acquiring building distribution data, wall material data and terrain elevation data according to the geographic topology data, and acquiring shielding attenuation coefficients and terrain attenuation coefficients in the signal propagation process according to the building distribution data, the wall material data and the terrain elevation data; acquiring a basic path loss parameter in the signal propagation process according to the network infrastructure layout data, and acquiring a propagation loss coefficient of the dual-mode network according to the basic path loss parameter, the shielding attenuation coefficients and the terrain attenuation coefficients; acquiring an interference loss coefficient according to the fixed interference source data and the dynamic interference source data, and acquiring a total loss influence factor according to the interference loss coefficient and the propagation loss coefficient; calibrating the initial signal strength according to the total loss influence factor to obtain the calibrated signal strength.

3. The heterogeneous network dual-mode adaptive handover method of claim 2, wherein, The step of acquiring the interference loss coefficient according to the fixed interference source data and the dynamic interference source data comprises: acquiring a fixed signal strength, a fixed center frequency and a fixed interference frequency bandwidth of each fixed interference source according to the fixed interference source data, and acquiring a corresponding fixed effective interference frequency bandwidth according to each of the fixed center frequency and the fixed interference frequency bandwidth; acquiring a corresponding fixed effective interference power according to each of the fixed effective interference frequency bandwidth, the fixed interference frequency bandwidth and the fixed signal strength, and acquiring a fixed total interference power according to a plurality of the fixed effective interference powers; According to the dynamic interference source data, dynamic signal strength, dynamic center frequency and dynamic interference frequency bandwidth of each dynamic interference source are obtained, and according to each dynamic center frequency and dynamic interference frequency bandwidth, corresponding dynamic effective interference frequency bandwidth is obtained; According to each dynamic effective interference frequency bandwidth, dynamic interference frequency bandwidth and dynamic signal strength, corresponding dynamic effective interference power is obtained, and according to a plurality of dynamic effective interference powers, dynamic total interference power is obtained; According to the fixed total interference power and dynamic total interference power, the comprehensive interference power of the dual-mode network is obtained; The useful signal strength of the dual-mode network is obtained, and according to the comprehensive interference power and the useful signal strength, the interference loss coefficient is obtained.

4. The heterogeneous network dual-mode adaptive handover method of claim 1, wherein, The step of obtaining the switching demand evaluation value according to the connection state characteristics and the side network consumption index comprises: obtaining connection state characteristics, wherein the connection state characteristics comprise connection stability sub-features, data transmission rate sub-features, connection time length sub-features and switching history success rate sub-features; obtaining a connection stability comprehensive value according to the connection stability sub-features, and obtaining a data transmission rate comprehensive value according to the data transmission rate sub-features; obtaining a connection time length comprehensive value according to the connection time length sub-features, and obtaining a switching history success rate comprehensive value according to the switching history success rate sub-features; performing weighted summation according to the connection stability comprehensive value, the data transmission rate comprehensive value, the connection time length comprehensive value and the switching history success rate comprehensive value to obtain a connection state comprehensive value; obtaining a side network consumption index, wherein the side network consumption index comprises unit data transmission power consumption sub-features, bandwidth utilization rate sub-features and battery residual capacity sub-features; obtaining a unit power consumption comprehensive value according to the unit data transmission power consumption sub-features, and obtaining a bandwidth utilization rate comprehensive value according to the bandwidth utilization rate sub-features; obtaining a battery capacity comprehensive value according to the battery residual capacity sub-features, and performing weighted summation according to the unit power consumption comprehensive value, the bandwidth utilization rate comprehensive value and the battery capacity comprehensive value to obtain a consumption index comprehensive value; obtaining an application demand weight coefficient of the heterogeneous network, and obtaining a switching demand evaluation value of the heterogeneous network according to the application demand weight coefficient of the connection state comprehensive value and the consumption index comprehensive value.

5. The heterogeneous network dual-mode adaptive handover method of claim 1, wherein, The step of generating a time decision sequence of the dual-mode network according to a plurality of signal fluctuation parameters and device power consumption change parameters comprises: obtaining signal fluctuation parameters, wherein the signal fluctuation parameters comprise signal strength change rate, transmission delay fluctuation amplitude and packet loss rate fluctuation amplitude; obtaining an environment scene type, and respectively performing scene-based weight distribution on each signal strength change rate, transmission delay fluctuation amplitude and packet loss rate fluctuation amplitude according to the environment scene type to obtain corresponding signal strength fluctuation weight, delay fluctuation weight and packet loss rate fluctuation weight; performing weighted summation according to each signal strength change rate, transmission delay fluctuation amplitude, packet loss rate fluctuation amplitude, signal strength fluctuation weight, delay fluctuation weight and packet loss rate fluctuation weight to obtain a corresponding signal fluctuation comprehensive evaluation value; obtaining device power consumption change parameters, wherein the device power consumption change parameters include instantaneous power consumption change rate, average power consumption change rate and peak power consumption change rate; performing scenario-based weight distribution on each of the instantaneous power consumption change rate, the average power consumption change rate and the peak power consumption change rate according to the environment scene type, to obtain corresponding instantaneous power consumption weight, average power consumption weight and peak power consumption weight; performing weighted summation according to each of the instantaneous power consumption change rate, the average power consumption change rate, the peak power consumption change rate, the instantaneous power consumption weight, the average power consumption weight and the peak power consumption weight, to obtain a corresponding power consumption change comprehensive evaluation value; obtaining a corresponding network comprehensive score according to each of the power consumption change comprehensive evaluation value and the signal fluctuation comprehensive evaluation value, and sorting multiple network comprehensive scores in order of size to obtain a time decision sequence.

6. The heterogeneous network dual-mode adaptive handover method of claim 1, wherein, The step of obtaining a corresponding directed time sequence graph according to multiple network state snapshot data, and obtaining a corresponding optimized time sequence graph according to each of the directed time sequence graph, comprises: obtaining network performance features and environment correlation features at a corresponding time according to each of the network state snapshot data, and constructing a time feature vector at a corresponding time according to each of the network performance features and environment correlation features; obtaining a corresponding feature similarity according to the time feature vectors of any two adjacent time points, and determining whether each of the feature similarities is less than a preset similarity; if the feature similarity is not less than the preset similarity, it is determined that the network state of the two adjacent time points corresponding to the feature similarity has no significant change, and a directed correlation edge between the two adjacent time points needs to be established; if the feature similarity is less than the preset similarity, it is determined that the network state of the two adjacent time points corresponding to the feature similarity has a significant change, and a directed correlation edge between the two adjacent time points does not need to be established; defining each time point in each time decision sequence as a graph node, and organizing multiple graph nodes and directed correlation edges in time sequence to form a corresponding directed time sequence graph; obtaining a local density of each node in each directed time sequence graph, and determining whether each of the local densities is less than a preset density; if the local density is less than the preset density, it is determined that the node is a redundant node, and each redundant node and its corresponding directed correlation edge are deleted from the directed time sequence graph to obtain a corresponding optimized time sequence graph.

7. A heterogeneous network dual-mode adaptive handover system, characterized by, comprises: a dynamic calibration module configured to obtain multi-dimensional environment perception data in a dual-mode coverage area of a heterogeneous network and an initial signal strength of the dual-mode network, and perform dynamic calibration on the initial signal strength according to the multi-dimensional environment perception data to obtain a calibrated signal strength; a first obtaining module configured to obtain a connection state feature and a side network consumption index of the dual-mode network under the calibrated signal strength, and obtain a handover demand evaluation value according to the connection state feature and the side network consumption index; a determination module configured to determine a handover demand level of the dual-mode network according to the handover demand evaluation value, and obtain multiple signal fluctuation parameters and device power consumption change parameters corresponding to the handover demand level; The first generation module is configured to generate time decision sequences of the dual-mode network according to the plurality of signal fluctuation parameters and the device power consumption change parameter, and obtain network state snapshot data of the dual-mode network at a plurality of continuous time instants in each time decision sequence; The second acquisition module is configured to obtain a corresponding directed time sequence graph according to a plurality of the network state snapshot data, and obtain a corresponding optimized time sequence graph according to each of the directed time sequence graphs; The second generation module is configured to obtain a corresponding network performance data set according to each of the time decision sequences, and generate a dual-mode adaptive switching model according to a plurality of the network performance data sets and the optimized time sequence graph, so that the device performs dual-mode network switching according to the dual-mode adaptive switching model.

8. The heterogeneous network dual-mode adaptive handover system of claim 7, wherein, The dynamic calibration module comprises: The first acquisition unit is configured to acquire static environment features and dynamic interference factors in the multi-dimensional environment perception data, wherein the static environment features comprise geographic topology data and network infrastructure layout data, and the dynamic interference factors comprise fixed interference source data and dynamic interference source data; The second acquisition unit is configured to acquire building distribution data, wall material data and terrain elevation data according to the geographic topology data, and acquire an occlusion attenuation coefficient and a terrain attenuation coefficient in a signal propagation process according to the building distribution data, the wall material data and the terrain elevation data; The third acquisition unit is configured to acquire a basic path loss parameter in a signal propagation process according to the network infrastructure layout data, and acquire a propagation loss coefficient of the heterogeneous network dual-mode according to the basic path loss parameter, the occlusion attenuation coefficient and the terrain attenuation coefficient; The fourth acquisition unit is configured to acquire an interference loss coefficient according to the fixed interference source data and the dynamic interference source data, and acquire a total loss influence factor according to the interference loss coefficient and the propagation loss coefficient; The calibration unit is configured to calibrate an initial signal strength according to the total loss influence factor to obtain a calibrated signal strength. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8. The processor executes the computer program to realize the steps of the method in any one of claims 1 to 6.

10. A computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to realize the steps of the method in any one of claims 1 to 6.

Images