System state prediction method and apparatus for mobile communication system

Abstract

The application relates to a system state prediction method for a mobile communication system, which comprises the following steps: acquiring system parameters of each subsystem in the communication system; constructing a corresponding subvector of each subsystem according to the system parameters of each subsystem; and constructing a system state vector of the communication system according to the subvectors; determining a clustering distance between the system state vector and a plurality of system state clustering results respectively; determining a target system state clustering result corresponding to the system state vector according to the clustering distances; and taking a system prediction state corresponding to the target system state clustering result as a target system prediction state. The method can reduce the calculation amount of system state prediction while retaining the accuracy of system state prediction.

Description

Technical Field

[0001] This application relates to the field of mobile communication technology, and in particular to a system state prediction method and apparatus for mobile communication systems. Background Technology

[0002] Mobile communication systems are an important component of communication systems. By integrating Integrated Services Digital Network (ISDN) and the Internet, mobile communication networks are organically combined with mobile communication networks, leveraging their respective strengths and compensating for each other. Furthermore, communication coverage can be extended through spaceborne platform communication systems and mobile satellite communication systems, thus forming an integrated communication system. Compared to civilian information and communication systems, mobile communication systems are characterized by a greater variety of equipment and terminals, and a more complex and variable operating environment.

[0003] Currently, most technologies for predicting the operational status of communication systems are focused on civilian communication systems. These systems are typically located in computer rooms, where the environment is relatively stable and the devices are far apart, so the mutual influence between them is negligible.

[0004] However, the operating environment of mobile communication systems is far more complex than that of civilian communication systems. Furthermore, the close proximity of devices in mobile communication systems means that the operation of one device can easily impact other nearby devices. If existing methods for predicting the operational status of civilian communication systems are used, numerous influencing factors must be considered, leading to complex modeling and high computational demands. Mobile communication systems may not be able to handle the computational burden required for system status prediction. Summary of the Invention

[0005] Therefore, it is necessary to provide a system state prediction method and apparatus for mobile communication systems to address the aforementioned technical problems.

[0006] Firstly, this application provides a system state prediction method for a mobile communication system. The method includes:

[0007] The system parameters of each subsystem in the communication system are obtained, and a subvector corresponding to each subsystem is constructed according to the system parameters of each subsystem. The system state vector of the communication system is constructed according to the subvectors. The system parameters include operating parameters and / or detected environmental parameters.

[0008] The clustering distances between the system state vector and multiple system state clustering results are determined respectively, and the target system state clustering result corresponding to the system state vector is determined according to each clustering distance. The system state clustering result includes at least one historical system state vector, which is a system state vector constructed in historical system state prediction.

[0009] The system prediction state corresponding to the clustering result of the target system state is taken as the target system prediction state.

[0010] In one embodiment, before obtaining the system parameters of each subsystem in the communication system, the method further includes:

[0011] Obtain each of the historical system state vectors, and the system prediction state corresponding to each of the historical system state vectors;

[0012] For any of the system prediction states, hierarchical clustering is performed on each of the historical system state vectors corresponding to the system prediction state to obtain multiple system state clustering results corresponding to the system prediction state.

[0013] In one embodiment, the hierarchical clustering process performed on the historical system state vectors corresponding to the predicted system state to obtain multiple system state clustering results corresponding to the predicted system state includes:

[0014] Each of the historical system state vectors is used as a temporary system state clustering result, and the clustering vector corresponding to each temporary system state clustering result is determined.

[0015] Determine the vector distance between every two cluster vectors in the cluster vectors, and take the smallest vector distance among all the vector distances as the target vector distance;

[0016] If the target vector distance is greater than the vector distance threshold, the clustering result of each temporary system state is taken as the clustering result of each system state, or;

[0017] If the target vector distance is less than or equal to the vector distance threshold, the temporary system state clustering results corresponding to the two clustering vectors corresponding to the target vector distance are merged into one temporary system state clustering result, and the process jumps to the step of determining the clustering vector corresponding to each temporary system state clustering result.

[0018] In one embodiment, determining the clustering distance between the system state vector and the clustering results of multiple system states includes:

[0019] For any clustering vector corresponding to the clustering result of the system state, the sub-vector distance between each sub-vector of the clustering vector and each sub-vector of the system state vector is determined respectively;

[0020] Based on the distances of each sub-vector, the vector distance between the clustering vector and the system state vector is determined, and the vector distance is used as the clustering distance between the system state vector and the system state clustering result.

[0021] In one embodiment, before obtaining the system parameters of each subsystem in the communication system, the method further includes:

[0022] In the case where a target subsystem exists in the communication system, and / or a target subvector exists in each of the historical system state vectors, the historical system state vectors are adjusted so that each subvector of the adjusted historical system state vector corresponds one-to-one with each of the subsystems in the communication system. The target subsystem is a subsystem that does not correspond to any of the subvectors, and the target subvector is a subvector that does not correspond to any of the subsystems.

[0023] The historical system state vectors are re-clustered to obtain multiple system state clustering results.

[0024] In one embodiment, when a target subsystem exists in the communication system, and / or a target subvector exists in each of the historical system state vectors, the adjustment of each of the historical system state vectors includes:

[0025] In the presence of a target subsystem, a subvector corresponding to the target subsystem is added to each of the historical system state vectors;

[0026] If a target subvector exists, the target subvector is deleted from each of the historical system state vectors.

[0027] In one embodiment, the method further includes:

[0028] Upon receiving a state adjustment instruction, the system predicted state corresponding to the system state vector is adjusted to the system predicted state indicated by the state adjustment instruction;

[0029] The system state vector after the system prediction state is adjusted is used as the historical system state vector.

[0030] Secondly, this application also provides a system state prediction device for a mobile communication system. The device includes:

[0031] A construction module is used to obtain system parameters of each subsystem in the communication system, construct sub-vectors corresponding to each subsystem according to the system parameters of each subsystem, and construct the system state vector of the communication system according to each sub-vector, wherein the system parameters include operating parameters and / or detected environmental parameters;

[0032] The determination module is used to determine the clustering distance between the system state vector and multiple system state clustering results, and to determine the target system state clustering result corresponding to the system state vector based on each clustering distance, wherein the system state clustering result includes at least one historical system state vector, and the historical system state vector is a system state vector constructed in historical system state prediction;

[0033] The first processing module is used to take the system prediction state corresponding to the clustering result of the target system state as the target system prediction state.

[0034] In one embodiment, the device further includes:

[0035] The acquisition module is used to acquire each of the historical system state vectors and the system prediction state corresponding to each of the historical system state vectors;

[0036] The first clustering module is used to perform hierarchical clustering processing on each of the historical system state vectors corresponding to any of the system prediction states, to obtain multiple system state clustering results corresponding to the system prediction states.

[0037] In one embodiment, the first clustering module is further configured to:

[0038] Each of the historical system state vectors is used as a temporary system state clustering result, and the clustering vector corresponding to each temporary system state clustering result is determined.

[0039] Determine the vector distance between every two cluster vectors in the cluster vectors, and take the smallest vector distance among all the vector distances as the target vector distance;

[0040] If the target vector distance is greater than the vector distance threshold, the clustering result of each temporary system state is taken as the clustering result of each system state, or;

[0041] If the target vector distance is less than or equal to the vector distance threshold, the temporary system state clustering results corresponding to the two clustering vectors corresponding to the target vector distance are merged into one temporary system state clustering result, and the process jumps to the step of determining the clustering vector corresponding to each temporary system state clustering result.

[0042] In one embodiment, the determining module is further configured to:

[0043] For any clustering vector corresponding to the clustering result of the system state, the sub-vector distance between each sub-vector of the clustering vector and each sub-vector of the system state vector is determined respectively;

[0044] Based on the distances of each sub-vector, the vector distance between the clustering vector and the system state vector is determined, and the vector distance is used as the clustering distance between the system state vector and the system state clustering result.

[0045] In one embodiment, the device further includes:

[0046] The first adjustment module is used to adjust each of the historical system state vectors when there is a target subsystem in the communication system and / or a target subvector in each of the historical system state vectors, so that each of the subvectors of the adjusted historical system state vectors corresponds one-to-one with each of the subsystems in the communication system, wherein the target subsystem is a subsystem that does not have a corresponding relationship with any of the subvectors, and the target subvector is a subvector that does not have a corresponding relationship with any of the subsystems;

[0047] The second clustering module is used to re-cluster each of the historical system state vectors to obtain multiple system state clustering results.

[0048] In one embodiment, the first adjustment module is further configured to:

[0049] In the presence of a target subsystem, a subvector corresponding to the target subsystem is added to each of the historical system state vectors;

[0050] If a target subvector exists, the target subvector is deleted from each of the historical system state vectors.

[0051] In one embodiment, the device further includes:

[0052] The second adjustment module is used to adjust the system predicted state corresponding to the system state vector to the system predicted state indicated by the state adjustment instruction when a state adjustment instruction is received.

[0053] The second processing module is used to take the system state vector after the system prediction state is adjusted as the historical system state vector.

[0054] Thirdly, this application also provides a computer device. The computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement any of the methods described above.

[0055] Fourthly, this application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements any of the above methods.

[0056] Fifthly, this application also provides a computer program product. The computer program product includes a computer program that, when executed by a processor, implements any of the above methods.

[0057] The aforementioned system state prediction method and apparatus for mobile communication systems constructs sub-vectors based on the system parameters (including operating parameters and environmental parameters) of each subsystem in the communication system. Then, it obtains the system state vector of the communication system based on each sub-vector, and uses a clustering algorithm to determine the clustering result of the system state vectors to which they belong, thereby obtaining the predicted state of the target system corresponding to the system state vectors. This application embodiment models the mobile communication system as a whole using vector modeling, considering all subsystems and their operating parameters, as well as detected environmental parameters, within the vector. By processing the communication system as a vector and incorporating environmental parameters into the calculation of distances between system states, the influence of the environment and the mutual influence between subsystems can be taken into account when calculating distances. This approach reduces computational load by simplifying the modeling of the communication system while maintaining the accuracy of system state prediction. Attached Figure Description

[0058] Figure 1 This is a flowchart illustrating a system state prediction method for a mobile communication system in one embodiment.

[0059] Figure 2 This is a flowchart illustrating a system state prediction method for a mobile communication system in one embodiment.

[0060] Figure 3 This is a flowchart illustrating step 204 in one embodiment;

[0061] Figure 4 This is a flowchart illustrating step 104 in one embodiment;

[0062] Figure 5 This is a flowchart illustrating a system state prediction method for a mobile communication system in one embodiment.

[0063] Figure 6 This is a flowchart illustrating step 502 in one embodiment;

[0064] Figure 7 This is a flowchart illustrating a system state prediction method for a mobile communication system in one embodiment.

[0065] Figure 8 This is a schematic diagram illustrating the changing trends of false alarm rate and false negative rate in one embodiment;

[0066] Figure 9This is a structural block diagram of a system state prediction device for a mobile communication system in one embodiment;

[0067] Figure 10 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0068] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0069] In one embodiment, such as Figure 1 As shown, a system state prediction method for a mobile communication system is provided. This embodiment illustrates the application of this method to a mobile communication vehicle within a mobile communication system, and includes the following steps:

[0070] Step 102: Obtain the system parameters of each subsystem in the communication system, construct the sub-vectors corresponding to each subsystem based on the system parameters of each subsystem, and construct the system state vector of the communication system based on each sub-vector. The system parameters include operating parameters and / or detected environmental parameters.

[0071] In this embodiment, the communication system is a communication system mounted on a mobile communication vehicle. The communication system may include multiple subsystems, such as a sub-communication system for external communication, a security system for encrypting information, and an environmental system for detecting environmental parameters such as temperature and humidity around the mobile communication vehicle. System parameters refer to the operating parameters (such as voltage, power, data transmission rate, etc.) generated by each subsystem during operation, as well as the environmental parameters (such as temperature, humidity, atmospheric pressure, etc.) detected when monitoring the external environment.

[0072] For any subsystem, a subvector corresponding to that subsystem can be constructed based on its system parameters. The subvector can be constructed based on all system parameters of the subsystem, or a subset of parameters (e.g., parameters closely related to the system state) can be selected from all system parameters. This application does not specifically limit this approach. With i1, i2, i3...i m Taking a system parameter as an example, the subsystem The corresponding subvector can be written as:

[0073] Concatenating the sub-vectors yields the system state vector corresponding to the communication system. If the communication system has multiple subsystems... The system state vector corresponding to the communication system can then be written as: Since some subsystems may have a greater impact on the operational state of the communication system, a component representing the weight of the subsystem can be added to the subvector. For example, if the subsystem... The weight is When constructing subvectors, each system parameter can be multiplied by its weight, i.e., the subsystem... The corresponding subvector can be The weights can be set by those skilled in the art according to actual needs. For example, when a subsystem is more important, the weight corresponding to the subsystem can be set to be larger, and when the importance of a subsystem is relatively weak, the weight corresponding to the subsystem can be set to be smaller.

[0074] Meanwhile, a communication system may contain more than one subsystem of the same type. For example, a single communication system may have multiple sub-communication systems used for simultaneous external communication. In this case, subvectors can be constructed for each subsystem of the same type, or a single subvector can be used to represent multiple subsystems of the same type. When representing multiple subsystems of the same type using a single subvector, the vectors corresponding to each subsystem can be constructed first, and then the vectors can be concatenated to obtain the subvector. For example, if there are subsystems... and according to The vector constructed from the system parameters is according to The vector constructed from the system parameters is Then you can The corresponding vector sum Concatenate the corresponding vectors to form sub-vectors.

[0075] Mathematically, this process is equivalent to mapping the system parameters in a communication system from a one-dimensional vehicle-mounted equipment space X to a high-dimensional global feature space F, where F is a Euclidean space. Let the mapping function be Φ: X → F, then the communication system x... i The mapping in F is

[0076] Step 104: Determine the clustering distances between the system state vector and multiple system state clustering results, and determine the target system state clustering result corresponding to the system state vector based on each clustering distance. The system state clustering result includes at least one historical system state vector, which is a system state vector constructed in the historical system state prediction.

[0077] In this embodiment, the system state clustering result is obtained by clustering multiple historical system state vectors, where the historical system state vectors refer to the system state vectors constructed in the historical system state prediction. For any system state clustering result, the clustering distance between the system state vector and the system state clustering result can be obtained based on the distance between the system state vector and each historical system state vector in the system state clustering result. This embodiment does not specifically limit the method for calculating the distance between the system state vector and the historical system state vectors; any method that can calculate the distance between two vectors is applicable to this embodiment, such as Euclidean distance, Bach distance, etc.

[0078] After obtaining the vector distances between the system state vector and each historical system state vector, the clustering distance between the system state vector and the clustering results of each system state can be determined based on the vector distances. This application does not specifically limit the method of obtaining the clustering distance from the vector distances. For example, the sum of the vector distances can be used as the clustering distance, or the average of the vector distances can be used as the clustering distance, etc. This application does not specifically limit this method.

[0079] After determining the clustering distance between the system state vector and the clustering results of each system state, the clustering result of the system state with the smallest corresponding clustering distance can be used as the target system state clustering result.

[0080] Step 106: Take the system prediction state corresponding to the clustering result of the target system state as the target system prediction state.

[0081] In this embodiment, each system state clustering result has a corresponding predicted system state, such as normal state or abnormal state. Alternatively, a finer-grained division can be made, such as good state, critical state, or abnormal state. This embodiment does not specifically limit this division.

[0082] The system predicted state corresponding to the system state clustering result can be determined based on the system predicted state corresponding to each historical system state vector in the system state clustering result. The system predicted state corresponding to the historical system state vector is obtained from the historical system state prediction. For example, if a system state clustering result includes historical system state vectors corresponding to multiple system predicted states, the system predicted state with the largest number of corresponding historical system state vectors can be selected as the system predicted state corresponding to the system state clustering result. Alternatively, other methods can be used to determine the system predicted state corresponding to the system state clustering result. This application embodiment does not specifically limit this method.

[0083] Since the clustering result of the target system state is the clustering result closest to the system state vector, the predicted system state corresponding to the system state vector is highly likely to be the same as the predicted system state corresponding to the target system state clustering result. Therefore, the predicted system state corresponding to the target system state clustering result can be used as the predicted system state corresponding to the system state vector. After obtaining the predicted system state corresponding to the system state vector, the predicted system state can be displayed on the display device of the communication system. When the predicted system state indicates an anomaly in the communication system, an alarm can also be issued in the form of sound, light, etc. This application embodiment does not specifically limit this.

[0084] The system state prediction method for mobile communication systems provided in this application constructs sub-vectors based on the system parameters (including operating parameters and environmental parameters) of each subsystem in the communication system. Then, it obtains the system state vector of the communication system based on each sub-vector, and uses a clustering algorithm to determine the clustering result of the system state vectors to which they belong, thereby obtaining the predicted state of the target system corresponding to the system state vectors. This application model the mobile communication system as a whole using vector modeling, considering all subsystems and their operating parameters, as well as detected environmental parameters, within the vector. By processing the communication system as a vector and including environmental parameters in the calculation of distances between system states, the method takes into account the influence of the environment and the mutual influence between subsystems when calculating distances. This approach maintains the accuracy of system state prediction while simplifying the modeling of the communication system and reducing computational load.

[0085] In one embodiment, such as Figure 2 As shown, in step 102, before obtaining the system parameters of each subsystem in the communication system, the above method further includes:

[0086] Step 202: Obtain the state vector of each historical system and the predicted state of the system corresponding to each historical system state vector.

[0087] Step 204: For any system prediction state, perform hierarchical clustering on the historical system state vectors corresponding to the system prediction state to obtain multiple system state clustering results corresponding to the system prediction state.

[0088] In this embodiment, after each system state prediction, the system state vector constructed in this prediction and the predicted system state based on the system state vector can be stored. When performing the next system state prediction, all previously stored system state vectors can be used as historical system state vectors and clustered once; that is, before each system state prediction begins, the previous prediction results are used as samples for re-clustering. Alternatively, when a certain number of system state vectors are stored, the stored system state vectors can be used as historical system state vectors and clustered together with the original historical system state vectors; that is, clustering is performed again after a fixed number of system state predictions. This embodiment does not specifically limit this approach.

[0089] When the number of historical system state vectors is insufficient to support automatic system state prediction, historical system state vectors can be constructed by acquiring stored system state data. System state data consists of system parameters stored at a specific moment during the historical operation of the communication system. However, since not all system state data is used for system state prediction, this portion of system state data has not yet been constructed into historical system state vectors, and there are no corresponding predicted system states. The system state data can be constructed into historical system state vectors according to the method described in the previous embodiments, thereby obtaining the system operating state of the communication system at the time the system state data was stored, and using this operating state as the predicted system state corresponding to the historical system state vector. Mathematically, this process is equivalent to defining the historical system state vector corresponding to a certain predicted system state as... in satisfy by For example, satisfy ( Defined as system parameters generated under the predicted state of the system.

[0090] Since this application embodiment only uses the historical system state vectors from a single mobile communication vehicle for clustering, the sample size is small. Therefore, hierarchical clustering is suitable for clustering the historical system state vectors. Hierarchical clustering involves merging the two closest historical system state vectors during each clustering process until the number of clusters reaches a preset number, or the distance between any two historical state vectors is greater than a distance threshold. The historical system state vectors corresponding to each system prediction state can be clustered separately. That is, if there are system prediction states A, B, and C, the historical system state vectors corresponding to state A, B, and C can be clustered separately, resulting in multiple system state clustering results for system prediction states A, B, and C. Each system state clustering result can represent a scenario under a system prediction state. For example, if there are 10 system state clustering results for the system prediction state of "normal," then each system state clustering result represents 10 categories of system parameter values ​​that can be interpreted as a normal system state.

[0091] When determining the clustering distance between the system state vector and the system state clustering results, the clustering distance between the system state vector and each system state clustering result can also be determined separately for the system state clustering results under each system prediction state.

[0092] The system state prediction method for mobile communication systems provided in this application uses historical system state vectors obtained from historical system state predictions as samples. Hierarchical clustering is then performed on each historical system state vector according to different predicted system states to obtain system state clustering results. This application uses historical system state vectors of mobile communication vehicles as samples, allowing the mobile communication vehicle to obtain system state clustering results without communicating with other devices, using its own historical data for training, thus reducing communication resource consumption. Furthermore, using hierarchical clustering, a low-resource clustering method, to cluster historical system state vectors further reduces the resource consumption of the mobile communication vehicle during clustering and achieves good clustering results even with a small number of samples.

[0093] In one embodiment, such as Figure 3 As shown, in step 204, hierarchical clustering is performed on the historical system state vectors corresponding to the predicted system state to obtain multiple system state clustering results corresponding to the predicted system state, including:

[0094] Step 302: Use each historical system state vector as a temporary system state clustering result, and determine the clustering vector corresponding to each temporary system state clustering result.

[0095] Step 304: Determine the vector distance between every two cluster vectors in the cluster vector, and take the smallest vector distance as the target vector distance.

[0096] Step 306: If the target vector distance is greater than the vector distance threshold, the clustering result of each temporary system state is taken as the clustering result of each system state, or;

[0097] Step 308: If the target vector distance is less than or equal to the vector distance threshold, merge the temporary system state clustering results corresponding to the two clustering vectors corresponding to the target vector distance into one temporary system state clustering result, and jump to the step of determining the clustering vectors corresponding to each temporary system state clustering result.

[0098] In this embodiment of the application, when performing hierarchical clustering on each historical system state vector, each historical system state vector can first be used as a temporary system state clustering result. That is, at the beginning of the clustering, each temporary system state clustering result contains one historical system state vector. At this time, for any temporary system state clustering result, the clustering vector corresponding to the temporary system state clustering result is the historical system state vector in that temporary system state clustering result.

[0099] For each temporary system state clustering result under each system prediction state, the vector distance between any two clustering vectors under that system prediction state can be determined. This application does not specifically limit the method for calculating the distance between two clustering vectors; any method that can calculate the distance between two vectors is applicable to this application, such as Euclidean distance, Bach distance, etc.

[0100] After obtaining the vector distance between any two clustering vectors, the smallest vector distance among them is taken as the target vector distance. The temporary system state clustering results corresponding to the two clustering vectors with the target vector distance are merged into a single temporary system state clustering result. This process of determining the distance between the clustering vectors of any two temporary system state clustering results is repeated until the target vector distance is greater than a vector distance threshold. The vector distance threshold can be determined by those skilled in the art based on actual needs.

[0101] When there are more than one historical system state vector in a temporary system state clustering result, the average value of each historical system state vector can be taken as the clustering vector corresponding to the temporary system state clustering result. Alternatively, the sum of distances between each historical system state vector and other historical system state vectors can be calculated separately, and the historical system state vector with the smallest sum of distances can be taken as the clustering vector. This application does not make specific limitations on this.

[0102] The system state prediction method for mobile communication systems provided in this application uses hierarchical clustering to perform hierarchical clustering on each historical system state vector to obtain system state clustering results. By using hierarchical clustering, a low-resource clustering method, to cluster historical system state vectors, the resource consumption of mobile communication vehicles during clustering can be reduced, and better clustering results can be achieved when the number of samples is small.

[0103] In one embodiment, such as Figure 4 As shown, in step 104, the clustering distances between the system state vector and multiple system state clustering results are determined, including:

[0104] Step 402: For any clustering vector corresponding to the clustering result of the system state, determine the sub-vector distance between each sub-vector of the clustering vector and each sub-vector of the system state vector.

[0105] Step 404: Determine the vector distance between the clustering vector and the system state vector based on the distance between each sub-vector, and use the vector distance as the clustering distance between the system state vector and the system state clustering result.

[0106] In this embodiment, the vector distance between the system state vector and the clustering vector corresponding to the system state clustering result can be used as the clustering distance between the system state vector and the system state clustering result. For any sub-vector in the clustering vector, the sub-vector corresponding to that sub-vector can be determined from the system state vector (when two sub-vectors correspond to the same subsystem, the two sub-vectors correspond to each other), and the sub-vector distance between the two sub-vectors can be calculated; then, based on the sub-vector distance between each sub-vector, the vector distance between the clustering vector and the system state vector can be determined.

[0107] For example, for any two subvectors, the subvector distance can be determined by the differences between the system parameters in each subvector. When determining the subvector distance based on the differences between system parameters, a normalization factor for the system parameter can also be determined based on the largest and smallest system parameter among all historical system state vectors. That is, for system parameter 'a', the system parameter 'a' with the largest value can be determined from the system parameters 'a' in each historical system state vector. max The system parameter a with the smallest sum value min When calculating the difference between system parameters a1 and a2, a can be used as a... max and a min The difference between them is used as a normalization factor.

[0108] The distance between any two subvectors can be calculated using Bhattacharyya distance (see Formula (I)):

[0109]

[0110] in, Let m be any two corresponding subvectors. and The total number of system parameters, refer to The j-th system parameter in refer to The j-th system parameter in Refers to the sub-vectors of the state vectors of the entire historical system. The j-th system parameter o j Among them, the largest value is o j ; Refers to the sub-vectors of the state vectors of the entire historical system. The j-th system parameter o j Among them, the smallest value is o j .

[0111] If a subvector represents multiple subsystems of the same type, that is... The form is We can first determine the distance between each vector in the sub-vector, and then determine the sub-vector distance based on the distance between each vector (see formula (II)):

[0112]

[0113] in, refer to The k-th vector in refer to The k-th vector in refer to The j-th system parameter in refer to The j-th system parameter in the system.

[0114] The vector distance between the clustering vector and the system state vector can be determined based on the distances of each sub-vector. For example, the vector distance between the clustering vector and the system state vector can be obtained by summing, weighted summing, or taking the average of the distances of each sub-vector. This application does not impose specific limitations on this.

[0115] The system state prediction method for mobile communication systems provided in this application determines the distance between two vectors based on the distance between corresponding sub-vectors within the two vectors. This application also determines the differences between sub-vectors representing the same subsystem within the two vectors and determines the overall difference between the two vectors based on these differences. This approach maintains the accuracy of system state prediction while simplifying the modeling of the communication system to reduce computational load.

[0116] In one embodiment, such as Figure 5 As shown, in step 102, before obtaining the system parameters of each subsystem in the communication system, the above method further includes:

[0117] Step 502: In the case that there is a target subsystem in the communication system and / or a target subvector in each historical system state vector, adjust each historical system state vector so that each subvector of the adjusted historical system state vector corresponds one-to-one with each subsystem in the communication system. Here, the target subsystem is a subsystem that does not have a corresponding relationship with any subvector, and the target subvector is a subvector that does not have a corresponding relationship with any subsystem.

[0118] Step 504: Re-cluster the historical system state vectors to obtain multiple system state clustering results.

[0119] In this embodiment, the correspondence between each subsystem and subvector can be recorded in the subsystem (control system) responsible for predicting the system state in the communication system. Before acquiring the system parameters of each subsystem in the communication system each time, the control system can determine, based on the currently existing subsystems in the communication system and the subvectors in the historical system state vector (any historical system state vector is acceptable, because in this embodiment, the number of subvectors in each historical system state vector and the subsystems corresponding to the subvectors should be the same), the subsystems that do not correspond to any subvector in the historical system state vector (i.e., the subsystems added in the communication system after the last update of the historical system state vector (target subsystems)) and the subvectors that do not correspond to any subsystem (i.e., the subsystems removed in the communication system after the last update of the historical system state vector, and the subvector corresponding to this subsystem in the historical system state vector is the target subvector).

[0120] Because the clustering distance between the system state vector and the system state clustering result cannot accurately represent the relationship between them when a target subsystem or target subvector exists, it is necessary to adjust the historical system state vectors and re-cluster each historical system state vector. This application does not specifically limit the method of adjusting the historical system state vectors. For example, when a target subvector exists, it can be removed; when a target subsystem exists, a subvector representing the target subsystem can be added to the historical system state vectors; or, instead of adjusting the historical system state vectors, the subvector corresponding to the target subsystem is not constructed when constructing the system state vectors.

[0121] The system state prediction method for mobile communication systems provided in this application can adjust historical system state vectors and re-cluster each historical system state vector when a target subsystem or target subvector exists. Therefore, when adding or removing devices (i.e., subsystems) in the communication system, historical data collected by the original devices at different times can be used without rebuilding the sample library required for clustering, thus improving the availability of system state prediction.

[0122] In one embodiment, such as Figure 6 As shown, in step 502, when a target subsystem exists in the communication system and / or a target subvector exists in each historical system state vector, the historical system state vectors are adjusted, including:

[0123] Step 602: If a target subsystem exists, add a subvector corresponding to the target subsystem to each historical system state vector.

[0124] Step 604: If a target subvector exists, delete the target subvector from each historical system state vector.

[0125] In this embodiment, if a target subvector exists, it can be deleted from each historical system state vector. If a target subsystem exists, a subvector corresponding to the target subsystem can be added to each historical system state vector, so that the subvectors of each historical system state vector can correspond one-to-one with each current subsystem of the communication system.

[0126] When adding a subvector corresponding to the target subsystem, the system parameters of the target subsystem can be collected, the subvector can be constructed based on the system parameters of the target subsystem, and the constructed subvector can be added to the total historical system state vector. Alternatively, the system parameters in the subvector corresponding to the target subsystem can be obtained in other ways, such as using the standard operating parameters and standard environmental parameters of the target subsystem as the system parameters in the subvector. This application does not specifically limit this method.

[0127] The system state prediction method for mobile communication systems provided in this application can add a sub-vector corresponding to the target subsystem to each historical system state vector when a target subsystem exists, or delete the target sub-vector from each historical system state vector when a target sub-vector exists. Therefore, when adding or removing devices (i.e., subsystems) in the communication system, historical data collected by the original devices at different times can be used, without having to rebuild the sample library required for clustering, thus improving the availability of system state prediction.

[0128] In one embodiment, such as Figure 7 As shown, the above method also includes:

[0129] Step 702: Upon receiving a state adjustment instruction, adjust the system predicted state corresponding to the system state vector to the system predicted state indicated by the state adjustment instruction.

[0130] Step 704: Use the system state vector after the system prediction state adjustment as the historical system state vector.

[0131] In this embodiment, when maintenance personnel discover an error in the system's predicted state, they can manually adjust the system's predicted state. For example, in the event of a false alarm, maintenance personnel can adjust the final system predicted state of the system state prediction that caused the false alarm to normal; that is, the system predicted state corresponding to the system state vector constructed in that system state prediction will also become normal. Similarly, in the event of a missed alarm, maintenance personnel can also adjust the system predicted state to abnormal. Maintenance personnel can also adjust the system predicted state accordingly in other abnormal situations; this embodiment does not specifically limit this adjustment.

[0132] After the system prediction state is adjusted, the system state vector can be used normally as the historical system state vector, and clustering is performed periodically as described in the aforementioned embodiments, which will not be repeated here. Through this semi-supervised learning method, the false alarm rate and false negative rate can be significantly reduced after several adjustments to the erroneous system prediction state. The effect can be seen in [reference needed]. Figure 8 As shown.

[0133] The system state prediction method for mobile communication systems provided in this application allows maintenance personnel to manually adjust the predicted system state to avoid incorrect samples entering the cluster and causing clustering errors, thereby improving the accuracy of system state prediction.

[0134] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0135] Based on the same inventive concept, this application also provides a system state prediction device for a mobile communication system for implementing the system state prediction method for a mobile communication system described above. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations in one or more embodiments of the system state prediction device for a mobile communication system provided below can be found in the limitations of the system state prediction method for a mobile communication system described above, and will not be repeated here.

[0136] In one embodiment, such as Figure 9 As shown, a system state prediction device 900 for a mobile communication system is provided, comprising: a construction module 902, a determination module 904, and a first processing module 906, wherein:

[0137] The construction module 902 is used to obtain the system parameters of each subsystem in the communication system, construct the sub-vector corresponding to each subsystem according to the system parameters of each subsystem, and construct the system state vector of the communication system according to the sub-vectors. The system parameters include operating parameters and / or detected environmental parameters.

[0138] The determining module 904 is used to determine the clustering distance between the system state vector and multiple system state clustering results respectively, and to determine the target system state clustering result corresponding to the system state vector based on each clustering distance, wherein the system state clustering result includes at least one historical system state vector, and the historical system state vector is a system state vector constructed in historical system state prediction;

[0139] The first processing module 906 is used to take the system prediction state corresponding to the clustering result of the target system state as the target system prediction state.

[0140] In one embodiment, the device further includes:

[0141] The acquisition module is used to acquire each of the historical system state vectors and the system prediction state corresponding to each of the historical system state vectors;

[0142] The first clustering module is used to perform hierarchical clustering processing on each of the historical system state vectors corresponding to any of the system prediction states, to obtain multiple system state clustering results corresponding to the system prediction states.

[0143] The system state prediction device for mobile communication systems provided in this application constructs sub-vectors based on the system parameters (including operating parameters and environmental parameters) of each subsystem in the communication system. Then, it obtains the system state vector of the communication system based on each sub-vector, and uses a clustering algorithm to determine the clustering result of the system state vectors to which they belong, thereby obtaining the predicted state of the target system corresponding to the system state vectors. This application model the mobile communication system as a whole using vector modeling, considering all subsystems and their operating parameters, as well as detected environmental parameters, within the vector. By processing the communication system as a vector and including environmental parameters in the calculation of distances between system states, it can take into account the influence of the environment and the mutual influence between subsystems when calculating distances. While simplifying the modeling of the communication system and reducing computational load, it still maintains the accuracy of system state prediction.

[0144] In one embodiment, the first clustering module is further configured to:

[0145] Each of the historical system state vectors is used as a temporary system state clustering result, and the clustering vector corresponding to each temporary system state clustering result is determined.

[0146] Determine the vector distance between every two cluster vectors in the cluster vectors, and take the smallest vector distance among all the vector distances as the target vector distance;

[0147] If the target vector distance is greater than the vector distance threshold, the clustering result of each temporary system state is taken as the clustering result of each system state, or;

[0148] If the target vector distance is less than or equal to the vector distance threshold, the temporary system state clustering results corresponding to the two clustering vectors corresponding to the target vector distance are merged into one temporary system state clustering result, and the process jumps to the step of determining the clustering vector corresponding to each temporary system state clustering result.

[0149] In one embodiment, the determining module 904 is further configured to:

[0150] For any clustering vector corresponding to the clustering result of the system state, the sub-vector distance between each sub-vector of the clustering vector and each sub-vector of the system state vector is determined respectively;

[0151] Based on the distances of each sub-vector, the vector distance between the clustering vector and the system state vector is determined, and the vector distance is used as the clustering distance between the system state vector and the system state clustering result.

[0152] In one embodiment, the device further includes:

[0153] The first adjustment module is used to adjust each of the historical system state vectors when there is a target subsystem in the communication system and / or a target subvector in each of the historical system state vectors, so that each of the subvectors of the adjusted historical system state vectors corresponds one-to-one with each of the subsystems in the communication system, wherein the target subsystem is a subsystem that does not have a corresponding relationship with any of the subvectors, and the target subvector is a subvector that does not have a corresponding relationship with any of the subsystems;

[0154] The second clustering module is used to re-cluster each of the historical system state vectors to obtain multiple system state clustering results.

[0155] In one embodiment, the first adjustment module is further configured to:

[0156] In the presence of a target subsystem, a subvector corresponding to the target subsystem is added to each of the historical system state vectors;

[0157] If a target subvector exists, the target subvector is deleted from each of the historical system state vectors.

[0158] In one embodiment, the device further includes:

[0159] The second adjustment module is used to adjust the system predicted state corresponding to the system state vector to the system predicted state indicated by the state adjustment instruction when a state adjustment instruction is received.

[0160] The second processing module is used to take the system state vector after the system prediction state is adjusted as the historical system state vector.

[0161] The modules in the aforementioned system state prediction device for mobile communication systems can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0162] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 10As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a system state prediction method for a mobile communication system.

[0163] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0164] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0165] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the steps in the above method embodiments.

[0166] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0167] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0168] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0169] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0170] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A system state prediction method for mobile communication systems, characterized in that, The method includes: Obtain the state vector of each historical system and the predicted state of the system corresponding to each historical system state vector; For any of the predicted system states, the historical system state vectors are used as temporary system state clustering results, and the clustering vectors corresponding to each temporary system state clustering result are determined. Determine the vector distance between every two cluster vectors in the cluster vectors, and take the smallest vector distance among all the vector distances as the target vector distance; If the target vector distance is greater than the vector distance threshold, each temporary system state clustering result is used as the system state clustering result; or if the target vector distance is less than or equal to the vector distance threshold, the temporary system state clustering results corresponding to the two clustering vectors corresponding to the target vector distance are merged into one temporary system state clustering result, and the process jumps to the step of determining the clustering vectors corresponding to each temporary system state clustering result. The system parameters of each subsystem in the communication system are obtained, and a subvector corresponding to each subsystem is constructed according to the system parameters of each subsystem. The system state vector of the communication system is constructed according to the subvectors. The system parameters include operating parameters and / or detected environmental parameters. For any clustering vector corresponding to the clustering result of the system state, the sub-vector distance between each sub-vector of the clustering vector and each sub-vector of the system state vector is determined respectively; Based on the distances of each sub-vector, the vector distance between the clustering vector and the system state vector is determined, and the vector distance is used as the clustering distance between the system state vector and the system state clustering result. Based on the clustering distances, the target system state clustering result corresponding to the system state vector is determined, wherein the system state clustering result includes at least one historical system state vector, and the historical system state vector is a system state vector constructed in historical system state prediction. The system prediction state corresponding to the clustering result of the target system state is taken as the target system prediction state.

2. The method according to claim 1, characterized in that, Before obtaining the system parameters of each subsystem in the communication system, the method further includes: In the case where a target subsystem exists in the communication system, and / or a target subvector exists in each of the historical system state vectors, the historical system state vectors are adjusted so that each subvector of the adjusted historical system state vector corresponds one-to-one with each of the subsystems in the communication system. The target subsystem is a subsystem that does not correspond to any of the subvectors, and the target subvector is a subvector that does not correspond to any of the subsystems. The historical system state vectors are re-clustered to obtain multiple system state clustering results.

3. The method according to claim 2, characterized in that, In the case where a target subsystem exists in the communication system, and / or a target subvector exists in each of the historical system state vectors, adjustments are made to each of the historical system state vectors, including: In the presence of a target subsystem, a subvector corresponding to the target subsystem is added to each of the historical system state vectors; If a target subvector exists, the target subvector is deleted from each of the historical system state vectors.

4. The method according to claim 1, characterized in that, The method further includes: Upon receiving a state adjustment instruction, the system predicted state corresponding to the system state vector is adjusted to the system predicted state indicated by the state adjustment instruction; The system state vector after the system prediction state is adjusted is used as the historical system state vector.

5. A system state prediction device for a mobile communication system, characterized in that, The device includes: The acquisition module is used to acquire each historical system state vector and the system prediction state corresponding to each historical system state vector; The first clustering module is used to, for any predicted system state, take each historical system state vector as a temporary system state clustering result and determine the clustering vector corresponding to each temporary system state clustering result; determine the vector distance between every two clustering vectors and take the smallest vector distance as the target vector distance; if the target vector distance is greater than the vector distance threshold, take each temporary system state clustering result as the system state clustering result; or if the target vector distance is less than or equal to the vector distance threshold, merge the temporary system state clustering results corresponding to the two clustering vectors corresponding to the target vector distance into one temporary system state clustering result, and jump to the step of determining the clustering vector corresponding to each temporary system state clustering result. A construction module is used to obtain system parameters of each subsystem in the communication system, construct sub-vectors corresponding to each subsystem according to the system parameters of each subsystem, and construct the system state vector of the communication system according to each sub-vector, wherein the system parameters include operating parameters and / or detected environmental parameters; The determining module is configured to, for any clustering vector corresponding to the system state clustering result, determine the sub-vector distance between each sub-vector of the clustering vector and each sub-vector of the system state vector; determine the vector distance between the clustering vector and the system state vector based on each sub-vector distance, and use the vector distance as the clustering distance between the system state vector and the system state clustering result; and determine the target system state clustering result corresponding to the system state vector based on each clustering distance, wherein the system state clustering result includes at least one historical system state vector, and the historical system state vector is a system state vector constructed in historical system state prediction; The first processing module is used to take the system prediction state corresponding to the clustering result of the target system state as the target system prediction state.

6. The apparatus according to claim 5, characterized in that, The device further includes: The first adjustment module is used to adjust each of the historical system state vectors when there is a target subsystem in the communication system and / or a target subvector in each of the historical system state vectors, so that each of the subvectors of the adjusted historical system state vectors corresponds one-to-one with each of the subsystems in the communication system, wherein the target subsystem is a subsystem that does not have a corresponding relationship with any of the subvectors, and the target subvector is a subvector that does not have a corresponding relationship with any of the subsystems; The second clustering module is used to re-cluster each of the historical system state vectors to obtain multiple system state clustering results.

7. The apparatus according to claim 6, characterized in that, The first adjustment module is further configured to: In the presence of a target subsystem, a subvector corresponding to the target subsystem is added to each of the historical system state vectors; If a target subvector exists, the target subvector is deleted from each of the historical system state vectors.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images