Artificial intelligence-based portable base station networking method and system

Abstract

The application provides an artificial intelligence-based portable base station networking method and system, relates to the technical field of communication, and can acquire the obstacle density of the area where the portable base station is located, the distance between the portable base station and neighbor base stations and the signal interference ratio; inputs the obstacle density, the distance between the portable base station and neighbor base stations and the signal interference ratio into an environment perception model to obtain the link stability score corresponding to the portable base station; inputs the link stability score, the energy information of the portable base station and the device communication state and the energy information of the neighbor base station into a resilience topology model to obtain the topology evaluation information of the portable base station; the resilience weight vector at least includes first information for indicating the topology reconstruction priority and second information for indicating the connection type; and in the case where the first information is greater than a preset reconstruction priority threshold value, the networking operation is performed according to the second information. The connection demand of the portable base station in a dynamic environment can be met.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a portable base station networking method and system based on artificial intelligence. Background Technology

[0002] With the rapid development of wireless communication technology, portable base stations are increasingly used in dynamic scenarios, such as vehicle-mounted communication networks, temporary coverage for large-scale events, and communication support at disaster relief sites. Traditional wireless communication networking strategies are mostly based on static or quasi-static environment design, mainly focusing on link configuration between fixed base stations or resource allocation for single service types, which is difficult to adapt to the connection needs in dynamic environments. Summary of the Invention

[0003] The technical problem this application aims to solve is to provide a portable base station networking method and system based on artificial intelligence, capable of meeting the connectivity needs of portable base stations in dynamic environments. The specific solution is as follows:

[0004] A portable base station networking method based on artificial intelligence, applied to portable base stations, the method comprising:

[0005] Once the portable base station has been deployed and the presence of a neighboring base station has been detected, the obstacle density in the area where the portable base station is located, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio are obtained; the obstacle density is the percentage of obstacle volume per unit volume.

[0006] The obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio are input into a pre-built environmental perception model to obtain the link stability score corresponding to the portable base station.

[0007] The link stability score, the energy information and device communication status of the portable base station, and the energy information of the neighboring base stations are input into a pre-built resilient topology model to obtain the topology evaluation information of the portable base station; the topology evaluation information includes at least first information for indicating the priority of topology reconfiguration and second information for indicating the connection type.

[0008] If the first information is greater than the preset reconstruction priority threshold, the networking operation is performed according to the second information.

[0009] Optionally, the process of obtaining the obstacle density in the area where the portable base station is located, as described above, includes:

[0010] Obtain point cloud data of the area where the portable base station is located;

[0011] The obstacle density of the area where the portable base station is located is obtained based on the point cloud data analysis.

[0012] Optionally, the environmental perception model described above is a machine learning model trained based on historical environmental data, comprising an input layer, a feature fusion layer, and an output layer.

[0013] The step of inputting the obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio into a pre-built environmental perception model to obtain the link stability score corresponding to the portable base station includes:

[0014] The obstacle density, distance, and signal-to-interference ratio are normalized to obtain a standardized input vector;

[0015] The standardized input vector is input into the input layer of the environment perception model, and multi-dimensional features are cross-fused through the feature fusion layer to extract the influence features of each parameter on the link stability.

[0016] The link stability score corresponding to the portable base station is obtained by calculating the influence features through linear regression of the output layer.

[0017] Optionally, in the above method, the step of performing the networking operation based on the second information includes:

[0018] When the second information indicates a dynamic environment connection type, the historical movement trajectory data of the portable base station and neighboring base stations are obtained; a Markov prediction model is constructed based on the historical movement trajectory data to predict the probability distribution of the relative position change of the portable base station and the neighboring base stations in the future; and networking operations are performed according to the probability distribution.

[0019] When the second information indicates a mixed service connection type, network operations are performed according to the current transmission service type of the portable base station and the priority weight of the current transmission service.

[0020] Optionally, in the above method, the step of performing network formation operation based on the probability distribution includes:

[0021] If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the first displacement velocity range is greater than the probability threshold, then a multi-frequency backup link pool is established and a real-time link switching mechanism is configured to complete the networking operation.

[0022] If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the second displacement velocity range is greater than the probability threshold, then an adaptive link optimization strategy is adopted to adjust the antenna beam coverage or transmission power and retain the basic backup link to complete the networking operation.

[0023] If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the third displacement velocity range is greater than the probability threshold, then a fixed high-gain directional link is established, beam energy is concentrated, and the minimum redundant link is retained to complete the networking operation.

[0024] A portable base station networking system based on artificial intelligence, applied to portable base stations, the system comprising:

[0025] The acquisition unit is used to acquire, when the portable base station has been deployed and the existence of a neighboring base station has been detected, the obstacle density of the area where the portable base station is located, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio; the obstacle density is the proportion of obstacle volume per unit volume;

[0026] The scoring unit is used to input the obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio into a pre-built environmental perception model to obtain the link stability score corresponding to the portable base station.

[0027] An execution unit is configured to input the link stability score, the energy information and device communication status of the portable base station, and the energy information of the neighboring base stations into a pre-constructed resilient topology model to obtain the topology evaluation information of the portable base station; the topology evaluation information includes at least first information for indicating the priority of topology reconstruction and second information for indicating the connection type;

[0028] The networking unit is used to perform networking operations based on the second information when the first information is greater than a preset reconstruction priority threshold.

[0029] Optionally, in the above-described system, the acquisition unit includes:

[0030] The first acquisition subunit is used to acquire point cloud data of the area where the portable base station is located.

[0031] The analysis subunit is used to analyze the point cloud data to obtain the obstacle density of the area where the portable base station is located.

[0032] Optionally, in the above system, the environment perception model is a machine learning model trained based on historical environment data, including an input layer, a feature fusion layer, and an output layer;

[0033] The scoring unit includes:

[0034] The processing module is used to normalize the obstacle density, distance, and signal-to-interference ratio to obtain a standardized input vector;

[0035] The extraction module is used to input the standardized input vector into the input layer of the environment perception model, and perform multi-dimensional feature cross-fusion through the feature fusion layer to extract the influence features of each parameter on the link stability.

[0036] The calculation module is used to calculate the link stability score corresponding to the portable base station by performing linear regression calculation on the influence features through the output layer.

[0037] Optionally, in the above system, the networking unit includes:

[0038] The first networking module is used to acquire historical movement trajectory data of the portable base station and neighboring base stations when the second information indicates a dynamic environment connection type; construct a Markov prediction model based on the historical movement trajectory data to predict the probability distribution of the relative position change of the portable base station and the neighboring base stations in the future; and perform networking operations according to the probability distribution.

[0039] The second networking module is used to perform networking operations based on the current transmission service type of the portable base station and the priority weight of the current transmission service when the second information indicates a mixed service connection type.

[0040] Optionally, in the above-described system, the first networking module includes:

[0041] The first network submodule is used to establish a multi-frequency backup link pool and configure a real-time link switching mechanism to complete the networking operation if the probability that the relative displacement speed between the portable base station and the neighboring base station in the probability distribution belongs to the first displacement speed range is greater than the probability threshold.

[0042] The second network submodule is used to adopt an adaptive link optimization strategy to adjust the antenna beam coverage or transmission power and retain basic backup links if the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution is greater than the probability threshold.

[0043] The third network submodule is used to establish a fixed high-gain directional link, concentrate beam energy, and retain the minimum redundant link to complete the networking operation if the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the third displacement velocity range is greater than the probability threshold.

[0044] Based on the above, this application provides a portable base station networking method and system based on artificial intelligence. When the portable base station is deployed and a neighboring base station is detected, the obstacle density of the area where the portable base station is located, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio are obtained. The obstacle density is the proportion of obstacle volume per unit volume. The obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio are input into a pre-built environmental perception model to obtain a link stability score corresponding to the portable base station. The link stability score, the energy information and device communication status of the portable base station, and the energy information of the neighboring base station are input into a pre-built resilient topology model to obtain topology evaluation information for the portable base station. The topology evaluation information includes at least first information indicating the topology reconfiguration priority and second information indicating the connection type. If the first information is greater than a preset reconfiguration priority threshold, networking operations are performed based on the second information. Applying the method provided in this application can meet the connection requirements of portable base stations in dynamic environments. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0046] Figure 1 A flowchart illustrating a portable base station networking method based on artificial intelligence, provided for this application;

[0047] Figure 2 A flowchart illustrating a process for obtaining obstacle density in the area where a portable base station is located, provided in this application;

[0048] Figure 3 A flowchart illustrating the process of obtaining a link stability score corresponding to a portable base station, as provided in this application;

[0049] Figure 4 This is a schematic diagram of a portable base station networking system based on artificial intelligence, provided for this application. Detailed Implementation

[0050] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present application.

[0051] In the present application, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the phrase "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0052] An embodiment of the present invention provides an artificial intelligence-based portable base station networking method, which is applied to an electronic device. The electronic device may be a portable base station. The method flowchart of the method is as Figure 1 shown and specifically includes:

[0053] S101: When the portable base station is deployed and it is detected that a neighbor base station exists, obtain the obstacle density in the area where the portable base station is located, the distance between the portable base station and the neighbor base station, and the signal-to-interference ratio; the obstacle density is the ratio of the volume of obstacles in a unit volume.

[0054] In this embodiment, a portable base station refers to a wireless communication node with mobility or temporary deployment characteristics, such as a vehicle-mounted base station, an emergency mobile base station, a base station supporting a wearable device, etc. A portable base station is different from a fixed base station and supports flexible adjustment of the coverage range and communication parameters. A neighbor base station refers to another base station that has communication coverage overlap or potential link connection requirements with the current portable base station, such as a fixed base station or other portable base stations.

[0055] Optionally, the obstacle density is the ratio of the volume of obstacles in a unit volume of space in the area where the portable base station is located, and can be used to quantify the degree of obstruction of the environment to the propagation of wireless signals.

[0056] In this embodiment, the distance between the portable base station and the neighbor base station refers to the spatial geometric straight-line distance between the two base stations, which can be calculated by the Euclidean distance formula after obtaining the coordinates of the two base stations through the global positioning system, or obtained through signal propagation time measurement technology, such as calculating the distance using the time difference of the round-trip of a pulse signal.

[0057] In this embodiment, the signal-to-interference ratio (SIR) refers to the ratio of the target signal power received by the portable base station receiver to the surrounding interference signal power, which can be used to evaluate the quality of the communication link.

[0058] Optionally, if the portable base station has completed physical installation, communication protocol initialization (such as registration with the core network), and operating parameter configuration (such as transmit power and frequency), and has entered normal working state, the deployment of the portable base station is determined to be complete.

[0059] In this embodiment, if a synchronization signal or beacon frame is detected by the receiver of the portable base station, it is determined that a neighboring base station is detected.

[0060] S102: Input the obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio into the pre-built environmental perception model to obtain the link stability score corresponding to the portable base station.

[0061] In this embodiment, the environment perception model refers to a machine learning model trained based on historical environment data. Its input consists of key parameters characterizing the wireless communication environment, including obstacle density, distance between base stations, and signal-to-interference ratio. The output is a numerical indicator that quantifies link stability, namely, the link stability score. This model can adopt architectures such as neural networks, support vector machines (SVM), or random forests. Specifically, it is trained using historical environment data containing labeled data of obstacle density, distance, signal-to-interference ratio, and corresponding actual link stability, and is used to learn the mapping relationship between environmental parameters and link stability.

[0062] Optionally, the link stability score is defined as a quantitative indicator of the ability of the communication link between the portable base station and the neighboring base station to maintain a reliable connection in a dynamic environment. The value range is [0,1]. The higher the score, the stronger the ability of the link to maintain stable communication under the influence of obstacles, distance changes or interference. The link stability score is obtained by comprehensive analysis and calculation of input parameters through the environmental perception model.

[0063] S103: Input the link stability score, the energy information and device communication status of the portable base station, and the energy information of the neighboring base stations into a pre-built resilient topology model to obtain the topology evaluation information of the portable base station; the topology evaluation information includes at least first information for indicating the priority of topology reconstruction and second information for indicating the connection type.

[0064] In this embodiment, the resilient topology model refers to a machine learning or rule-driven model trained based on historical topology data. Its inputs include multi-dimensional parameters such as link stability scores, base station energy information, and communication status, and its output is topology assessment information that quantitatively evaluates network topology robustness and adjustment needs. The resilient topology model aims to improve network anti-interference capabilities and adaptive reconstruction efficiency. It is trained using historical data or constructed using a rule base to learn the correlation between topology parameters and network resilience. The historical data includes link stability scores, base station energy consumption records, communication status logs, and validity annotations of corresponding topology adjustment strategies. The rule base may include mapping relationships between various variables and reconstruction priorities, connection types, etc.

[0065] Optionally, energy information refers to the remaining battery power or energy consumption rate of the portable base station or neighboring base stations, used to characterize the base station's continuous operating capability. Device communication status refers to the current communication load of the portable base station, the fault status of the transmit / receive module, and protocol compatibility (such as support for 5G NR / 4G LTE), used to assess the feasibility of the base station participating in network deployment. The communication load may include the percentage of bandwidth already used.

[0066] In this embodiment, the topology evaluation information is the output of the resilient topology model, which includes at least two types of key parameters:

[0067] The first piece of information is used to quantify the urgency of the current network topology requiring structural adjustments. These adjustments can include switching connecting base stations, adding or removing links, etc. The value range is [0,1], where a higher value indicates a more urgent need for reconstruction.

[0068] The second piece of information is used to indicate the type of connection strategy that should be adopted for the current network operation, such as dynamic environment connection type, mixed service connection type, etc., which can be defined by discrete tags or descriptive fields.

[0069] In this embodiment, the step of inputting link stability score, portable base station energy information and device communication status, and neighboring base station energy information into a pre-built resilient topology model to obtain topology evaluation information specifically includes:

[0070] First, the input data is integrated and preprocessed. The link stability score, portable base station energy information, device communication status and neighbor base station energy information are aligned and standardized to form a multi-dimensional input vector.

[0071] Secondly, feature extraction and evaluation calculations are performed based on the model architecture type. If the model is a machine learning architecture, the input vector is processed by a feature extraction layer to extract time-series and spatial features, and then the first and second information are calculated by a fully connected decision layer. If the model is a rule-driven architecture, the input vector is matched using a preset rule base (e.g., "when the link stability score is <0.5 and the remaining battery power of the portable base station is <20%, the first information = 0.8, and the second information = dynamic environment connection type") to directly output topology evaluation information. The feature extraction layer includes an LSTM layer or an attention mechanism layer; time-series features can include, for example, historical trends in link stability and predicted energy consumption curves; spatial features can include energy differences between the portable base station and neighboring base stations.

[0072] S104: If the first information is greater than the preset reconstruction priority threshold, perform networking operation according to the second information.

[0073] In this embodiment, the step of performing network formation operation based on the second information when the first information is greater than a preset reconstruction priority threshold is specifically implemented as follows:

[0074] First, in response to the topology evaluation information output by the resilient topology model, the current value of the first information is obtained and compared with a preset reconstruction priority threshold. If the current value of the first information is greater than the threshold, it is determined that the current topology needs to be reconstructed.

[0075] Secondly, based on the connection type indicated by the second information, such as dynamic environment connection type or mixed service connection type, the corresponding operation set is retrieved from the preset network operation strategy library. For example, the dynamic environment connection strategy library includes link switching and parameter adjustment operations, and the mixed service connection strategy library includes protocol adaptation and resource reallocation operations.

[0076] Next, based on the matching set of operations, the communication control module of the portable base station, such as the baseband processor or network management unit, is invoked to perform specific operations. For example, when the second information is a dynamic environment connection type, the receiver is prioritized to rescan the neighboring base station signals and select the target base station with a higher link stability score to establish a connection in order to perform a link switching operation. When the second information is a mixed service connection type, the traffic scheduling module is prioritized to allocate voice service traffic to low-latency links and data service traffic to high-bandwidth links in order to perform a resource reallocation operation.

[0077] Finally, after the networking operation is completed, the effect of the topology adjustment is verified by monitoring modules such as the bit error rate detector and the link packet loss rate statistics unit, and the results are fed back to the resilient topology model for subsequent model training or rule base updates, such as adjusting the reconstruction priority threshold or the mapping relationship of connection types.

[0078] In this embodiment, a collaborative mechanism of threshold judgment and connection type matching is used to realize adaptive dynamic adjustment of network topology, which effectively improves the anti-interference capability and continuous service capability of portable base station network in complex environments.

[0079] In one embodiment provided in this application, based on the above-described solution, optionally, the process of obtaining the obstacle density of the area where the portable base station is located is as follows: Figure 2 As shown, it includes:

[0080] S201: Obtain point cloud data of the area where the portable base station is located.

[0081] In this embodiment, point cloud data is a set of three-dimensional spatial points collected by sensors such as lidar, multi-view cameras or millimeter-wave radar. Each point contains coordinate information and optional attributes such as reflection intensity and color, which are used to describe the geometric features of objects in the region.

[0082] Optionally, point cloud data can be acquired in real time by a LiDAR sensor deployed on the portable base station, or generated by aerial photography of the target area using a multi-camera mounted on a drone, or by calling pre-stored historical 3D map data, which may include point cloud data from the city's BIM model.

[0083] S202: Obstacle density of the area where the portable base station is located is obtained based on the point cloud data analysis.

[0084] In this embodiment, outlier noise can be removed by statistical filtering or radius filtering algorithms, and data redundancy can be reduced by downsampling algorithms. Point cloud segmentation algorithms are used to identify and extract point sets that represent obstacles. For example, points with a height greater than a preset threshold are classified as building or vegetation obstacles. The target area is divided into several equal-area sub-regions, and the number or volume ratio of obstacle points in each sub-region is counted. The average value is taken as the obstacle density of the region. Alternatively, the density can be quantified by calculating the spatial distribution entropy value of obstacle points.

[0085] In one embodiment provided in this application, based on the above scheme, optionally, the environment perception model is a machine learning model trained based on historical environment data, including an input layer, a feature fusion layer and an output layer;

[0086] The process of inputting the obstacle density, the distance between the portable base station and the neighboring base stations, and the signal-to-interference ratio into a pre-built environmental perception model to obtain the link stability score corresponding to the portable base station is as follows: Figure 3 As shown, it includes:

[0087] S301: Normalize the obstacle density, distance, and signal-to-interference ratio to obtain a standardized input vector.

[0088] Normalization algorithms are applied to the input obstacle density, distance, and signal interference ratio to map each parameter to the standard interval of [0,1] or [-1,1], forming a standardized input vector containing three-dimensional features.

[0089] S302: Input the standardized input vector into the input layer of the environment perception model, and perform multi-dimensional feature cross-fusion through the feature fusion layer to extract the influence features of each parameter on the link stability.

[0090] The standardized input vector is input into the input layer of the environment perception model, and multi-dimensional features are cross-fused through the feature fusion layer to extract the influence features of each parameter on the link stability. For example, the interaction features of obstacle density and distance are extracted to reflect the synergistic effect of signal occlusion and propagation loss, and the interaction features of signal interference ratio and obstacle density reflect the enhancement mechanism of interference signals under occlusion environment.

[0091] S303: The influence features are calculated by linear regression of the output layer to obtain the link stability score corresponding to the portable base station.

[0092] In this embodiment, the influencing features are input to the output layer, and a weighted summation is performed using a linear regression model to output a link stability score.

[0093] In one embodiment provided in this application, based on the above-described scheme, optionally, the step of performing the networking operation according to the second information includes:

[0094] When the second information indicates a dynamic environment connection type, the historical movement trajectory data of the portable base station and neighboring base stations are obtained; a Markov prediction model is constructed based on the historical movement trajectory data to predict the probability distribution of the relative position change of the portable base station and the neighboring base stations in the future; and networking operations are performed according to the probability distribution.

[0095] When the second information indicates a mixed service connection type, network operations are performed according to the current transmission service type of the portable base station and the priority weight of the current transmission service.

[0096] In this embodiment, the second information is used to indicate the classification identifier of the network connection scenario type, specifically including dynamic environment connection type and mixed service connection type; dynamic environment connection type represents the connection scenario where the portable base station and neighboring base stations are in a highly mobile and frequently changing environment; mixed service connection type represents the connection scenario where the portable base station needs to handle multiple transmission services of different priorities at the same time.

[0097] Optionally, historical movement trajectory data is obtained by using GPS positioning or base station ranging technology to obtain the location coordinate sequence of the portable base station and neighboring base stations within a historical time period, such as the previous T time units, where T≥1. The Markov prediction model is a stochastic process model constructed based on Markov properties, and its state space is a discrete or continuous location interval.

[0098] In this embodiment, the transition probability matrix is ​​obtained from historical movement trajectory data and is used to describe the transition probability of the location state. The relative position change probability distribution is the probability density function or quality function of the relative position distance, azimuth angle, etc. between the portable base station and neighboring base stations within the future time t+Δt, Δt>0 output by the model.

[0099] Optionally, the transmission service types include real-time voice, video streams, data file transmission, and other service categories that need to be transmitted; the priority weight is a quantitative weight assigned to each service with a value range of [0,1], and the sum of the weights is 1, representing the priority of the service's demand for network connection quality, latency, bandwidth, etc.

[0100] In this embodiment, when the second information indicates a dynamic environment connection type: First, acquire historical movement trajectory data of the portable base station and neighboring base stations, such as the position coordinate sequence of the previous T time units; second, construct a Markov prediction model, including discretizing the position state into several intervals, such as dividing it into [0, 10m), [10m, 20m), etc. according to distance, statistically analyzing the state transition frequency in the historical trajectory, and calculating the state transition probability matrix P(i,j) = N(i,j) / N(i), where N(i,j) is the number of transitions from state i to j, and N(i) is the total number of transitions from state i; then, input the current position state into the model, and predict the probability distribution of relative position changes in the future time through matrix multiplication; finally, perform networking operations according to the probability distribution, such as prioritizing the establishment of connections with neighboring base stations that are highly likely to maintain a short distance in the future, thereby reducing the risk of signal attenuation.

[0101] In this embodiment, when the second information indicates a mixed service connection type: First, the current transmission service type of the portable base station is identified by service identifier or protocol parsing; second, the priority weights of each service are obtained, either pre-configured or dynamically calculated according to service quality requirements, such as real-time voice weight of 0.7 and data transmission weight of 0.3; finally, network operations are performed according to the priority weights, such as allocating dedicated bandwidth for high-priority services or selecting neighboring base stations with lower latency to ensure their transmission quality.

[0102] This method uses Markov models to quantitatively predict location changes in dynamic environments and combines this with differentiated resource allocation based on business priority weights, effectively improving the stability of network connections and the quality of business services in complex scenarios.

[0103] In one embodiment provided in this application, based on the above-described scheme, optionally, the step of performing network formation operation according to the probability distribution includes:

[0104] If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the first displacement velocity range is greater than the probability threshold, then a multi-frequency backup link pool is established and a real-time link switching mechanism is configured to complete the networking operation.

[0105] If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the second displacement velocity range is greater than the probability threshold, then an adaptive link optimization strategy is adopted to adjust the antenna beam coverage or transmission power and retain the basic backup link to complete the networking operation.

[0106] If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the third displacement velocity range is greater than the probability threshold, then a fixed high-gain directional link is established, beam energy is concentrated, and the minimum redundant link is retained to complete the networking operation.

[0107] In this embodiment, the probability distribution can be the statistical probability distribution of the relative displacement velocity between the portable base station and its neighboring base stations in the future time, which is output by the Markov prediction model and represents the probability of occurrence corresponding to different displacement velocity values.

[0108] Optionally, relative displacement speed refers to the rate at which the relative position of the portable base station and its neighboring base stations changes per unit time; the first displacement speed range, the second displacement speed range, and the third displacement speed range are speed intervals pre-divided according to the network connection stability requirements, such as the first range corresponding to high-speed movement [V1, +∞), the second range corresponding to medium-speed movement [V2, V1), and the third range corresponding to low-speed movement [0, V2), where V1>V2>0.

[0109] Optionally, the probability threshold is a pre-defined probability critical value range (0,1], such as 0.7, used to trigger specific networking strategies.

[0110] In this embodiment, the multi-frequency backup link pool refers to a set of backup link resources containing multiple different communication frequencies such as 2.4GHz, 5GHz, etc.; the real-time link switching mechanism is a control logic that dynamically selects the current optimal link based on real-time displacement speed, such as selecting the link with the highest signal strength.

[0111] Optionally, adaptive link optimization strategy refers to optimization method that dynamically adjusts link parameters according to displacement velocity; antenna beam coverage range is the spatial coverage area of ​​the antenna transmitted signal, such as beamwidth θ;

[0112] In this embodiment, the transmission power is the energy intensity of the signal transmission.

[0113] Optionally, basic backup links refer to the minimum number of backup links to ensure basic communication needs; fixed high-gain directional links refer to directional communication links with fixed beam direction and high signal gain, such as gain G≥20dBi; concentrated beam energy refers to focusing the antenna's transmitted energy in a specific direction to improve the signal strength of the target area; minimum redundancy links refer to the minimum number of redundant links required to maintain network connectivity, such as 1 link.

[0114] The specific implementation process of network formation based on probability distribution is as follows:

[0115] First, a Markov prediction model is used to obtain the probability distribution of the relative displacement velocity between the portable base station and its neighboring base stations in the future. Second, it is determined whether the probability of the relative displacement velocity belonging to the first, second, and third displacement velocity ranges in the probability distribution is greater than the probability threshold, and the corresponding networking operations are performed accordingly.

[0116] If the relative displacement velocity falls within the first displacement velocity range, i.e., the probability of a high-speed movement scenario is greater than the probability threshold, then the following operations are performed: establish a multi-frequency backup link pool containing multiple different frequencies, such as configuring 2.4GHz and 5GHz links simultaneously, and configure a real-time link switching mechanism, such as detecting the signal strength of each link every Δt time and switching to the current optimal link. By using multi-frequency redundancy and dynamic switching, the risk of link interruption caused by high-speed movement is reduced, and the networking operation is completed.

[0117] If the relative displacement velocity falls within the second displacement velocity range, meaning the probability of a medium-speed movement scenario is greater than the probability threshold, then an adaptive link optimization strategy is executed: the antenna beam coverage is adjusted according to the current displacement velocity, such as increasing the beamwidth θ to expand the coverage area or increasing the transmission power to P to compensate for signal attenuation, while retaining basic backup links, such as two low-priority links; while ensuring coverage, excessive resource consumption is avoided, and the networking operation is completed.

[0118] If the relative displacement velocity falls within the third displacement velocity range, meaning the probability of a low-speed movement or stationary scenario is greater than the probability threshold, then a fixed high-gain directional link is established. For example, the beam direction is fixed to the location of the neighboring base station, with a gain G = 25 dBi. The signal strength in the target direction is enhanced by concentrating beam energy, and only the minimum redundant link, such as one link, is retained. This reduces the occupation of redundant resources while meeting communication quality requirements, thus completing the networking operation.

[0119] This method quantifies the probability distribution of relative displacement velocity and configures link strategies differently for different mobile scenarios. It achieves redundancy protection in high-speed scenarios, dynamic optimization in medium-speed scenarios, and efficient resource utilization in low-speed scenarios, significantly improving the reliability and resource utilization of network connections in complex mobile environments. It has clear technical feasibility and scenario adaptability.

[0120] See Figure 4 This is a schematic diagram of a portable base station networking system based on artificial intelligence, provided in an embodiment of this application. The system is applied to a portable base station and includes:

[0121] The acquisition unit 401 is used to acquire, when the portable base station has been deployed and the existence of a neighboring base station has been detected, the obstacle density of the area where the portable base station is located, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio; the obstacle density is the proportion of obstacle volume per unit volume;

[0122] The scoring unit 402 is used to input the obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio into a pre-built environmental perception model to obtain the link stability score corresponding to the portable base station.

[0123] The execution unit 403 is used to input the link stability score, the energy information and device communication status of the portable base station, and the energy information of the neighboring base stations into a pre-built resilient topology model to obtain the topology evaluation information of the portable base station; the topology evaluation information includes at least first information for indicating the priority of topology reconstruction and second information for indicating the connection type.

[0124] The networking unit 404 is used to perform networking operations based on the second information when the first information is greater than a preset reconstruction priority threshold.

[0125] In one embodiment provided in this application, based on the above-described solution, optionally, the acquisition unit 401 includes:

[0126] The first acquisition subunit is used to acquire point cloud data of the area where the portable base station is located.

[0127] The analysis subunit is used to analyze the point cloud data to obtain the obstacle density of the area where the portable base station is located.

[0128] In one embodiment provided in this application, based on the above scheme, optionally, the environment perception model is a machine learning model trained based on historical environment data, including an input layer, a feature fusion layer and an output layer;

[0129] The scoring unit 402 includes:

[0130] The processing module is used to normalize the obstacle density, distance, and signal-to-interference ratio to obtain a standardized input vector;

[0131] The extraction module is used to input the standardized input vector into the input layer of the environment perception model, and perform multi-dimensional feature cross-fusion through the feature fusion layer to extract the influence features of each parameter on the link stability.

[0132] The calculation module is used to calculate the link stability score corresponding to the portable base station by performing linear regression calculation on the influence features through the output layer.

[0133] In one embodiment provided in this application, based on the above-described scheme, optionally, the networking unit 404 includes:

[0134] The first networking module is used to acquire historical movement trajectory data of the portable base station and neighboring base stations when the second information indicates a dynamic environment connection type; construct a Markov prediction model based on the historical movement trajectory data to predict the probability distribution of the relative position change of the portable base station and the neighboring base stations in the future; and perform networking operations according to the probability distribution.

[0135] The second networking module is used to perform networking operations based on the current transmission service type of the portable base station and the priority weight of the current transmission service when the second information indicates a mixed service connection type.

[0136] In one embodiment provided in this application, based on the above-described solution, optionally, the first networking module includes:

[0137] The first network submodule is used to establish a multi-frequency backup link pool and configure a real-time link switching mechanism to complete the networking operation if the probability that the relative displacement speed between the portable base station and the neighboring base station in the probability distribution belongs to the first displacement speed range is greater than the probability threshold.

[0138] The second network submodule is used to adopt an adaptive link optimization strategy to adjust the antenna beam coverage or transmission power and retain basic backup links if the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution is greater than the probability threshold.

[0139] The third network submodule is used to establish a fixed high-gain directional link, concentrate beam energy, and retain the minimum redundant link to complete the networking operation if the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the third displacement velocity range is greater than the probability threshold.

[0140] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0141] Finally, it should be noted that in this paper, relational terms such as first and second are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations.

[0142] For ease of description, the above system is described by dividing it into various functional units. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware.

[0143] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0144] The above provides a detailed description of a portable base station networking method based on artificial intelligence provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A portable base station networking method based on artificial intelligence, characterized in that, Applied to portable base stations, the method includes: Once the portable base station has been deployed and the presence of a neighboring base station has been detected, the obstacle density in the area where the portable base station is located, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio are obtained; the obstacle density is the percentage of obstacle volume per unit volume. The obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio are input into a pre-built environmental perception model to obtain the link stability score corresponding to the portable base station. The environmental perception model is a machine learning model trained based on historical environmental data, including an input layer, a feature fusion layer, and an output layer. The link stability score, the energy information and communication status of the portable base station, and the energy information of the neighboring base stations are input into a pre-constructed resilient topology model to obtain the topology evaluation information of the portable base station. The topology evaluation information includes at least first information indicating the priority of topology reconstruction and second information indicating the connection type. The link stability score is a quantitative indicator representing the ability of the communication link between the portable base station and the neighboring base stations to maintain reliable connection in a dynamic environment. The first information is used to quantitatively represent the urgency of the current network topology requiring structural adjustment. The second information is used to indicate the connection strategy type that should be adopted for the current network operation. The resilient topology model refers to a machine learning or rule-driven model trained based on historical topology data. If the first information is greater than the preset reconstruction priority threshold, the networking operation is performed according to the second information; The step of inputting the obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio into a pre-built environmental perception model to obtain the link stability score corresponding to the portable base station includes: The obstacle density, distance, and signal-to-interference ratio are normalized to obtain a standardized input vector; The standardized input vector is input into the input layer of the environment perception model, and multi-dimensional features are cross-fused through the feature fusion layer to extract the influence features of each parameter on the link stability. The link stability score corresponding to the portable base station is obtained by calculating the impact features through linear regression of the output layer. The network formation operation based on the second information includes: When the second information indicates a dynamic environment connection type, the historical movement trajectory data of the portable base station and neighboring base stations are obtained; a Markov prediction model is constructed based on the historical movement trajectory data to predict the probability distribution of the relative position change of the portable base station and the neighboring base stations in the future; and networking operations are performed according to the probability distribution. When the second information indicates a mixed service connection type, network operations are performed according to the current transmission service type of the portable base station and the priority weight of the current transmission service.

2. The method according to claim 1, characterized in that, The process of obtaining the obstacle density of the area where the portable base station is located includes: Obtain point cloud data of the area where the portable base station is located; The obstacle density of the area where the portable base station is located is obtained based on the point cloud data analysis.

3. The method according to claim 1, characterized in that, The network formation operation based on the probability distribution includes: If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the first displacement velocity range is greater than the probability threshold, then a multi-frequency backup link pool is established and a real-time link switching mechanism is configured to complete the networking operation. If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the second displacement velocity range is greater than the probability threshold, then an adaptive link optimization strategy is adopted to adjust the antenna beam coverage or transmission power and retain the basic backup link to complete the networking operation. If the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the third displacement velocity range is greater than the probability threshold, then a fixed high-gain directional link is established, beam energy is concentrated, and the minimum redundant link is retained to complete the networking operation.

4. A portable base station networking system based on artificial intelligence, characterized in that, The system, applied to portable base stations, includes: The acquisition unit is used to acquire, when the portable base station has been deployed and the existence of a neighboring base station has been detected, the obstacle density of the area where the portable base station is located, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio; the obstacle density is the proportion of obstacle volume per unit volume; The scoring unit is used to input the obstacle density, the distance between the portable base station and the neighboring base station, and the signal-to-interference ratio into a pre-built environmental perception model to obtain the link stability score corresponding to the portable base station. An execution unit is configured to input the link stability score, the energy information and communication status of the portable base station, and the energy information of the neighboring base stations into a pre-constructed resilient topology model to obtain topology evaluation information for the portable base station. The topology evaluation information includes at least first information indicating the priority of topology reconfiguration and second information indicating the connection type. The link stability score is a quantitative indicator representing the ability of the communication link between the portable base station and the neighboring base stations to maintain reliable connectivity in a dynamic environment. The first information is used to quantitatively characterize the urgency of the current network topology requiring structural adjustments. The second information is used to indicate the connection strategy type to be adopted in the current network operation. The resilient topology model refers to a machine learning or rule-driven model trained based on historical topology data. A networking unit is used to perform networking operations based on the second information when the first information is greater than a preset reconstruction priority threshold. The environmental perception model is a machine learning model trained based on historical environmental data, including an input layer, a feature fusion layer, and an output layer. The scoring unit includes: The processing module is used to normalize the obstacle density, distance, and signal-to-interference ratio to obtain a standardized input vector; The extraction module is used to input the standardized input vector into the input layer of the environment perception model, and perform multi-dimensional feature cross-fusion through the feature fusion layer to extract the influence features of each parameter on the link stability. The calculation module is used to calculate the link stability score corresponding to the portable base station by performing linear regression calculation on the influence features through the output layer. The networking unit includes: The first networking module is used to acquire historical movement trajectory data of the portable base station and neighboring base stations when the second information indicates a dynamic environment connection type; construct a Markov prediction model based on the historical movement trajectory data to predict the probability distribution of the relative position change of the portable base station and the neighboring base stations in the future; and perform networking operations according to the probability distribution. The second networking module is used to perform networking operations based on the current transmission service type of the portable base station and the priority weight of the current transmission service when the second information indicates a mixed service connection type.

5. The system according to claim 4, characterized in that, The acquisition unit includes: The first acquisition subunit is used to acquire point cloud data of the area where the portable base station is located. The analysis subunit is used to analyze the point cloud data to obtain the obstacle density of the area where the portable base station is located.

6. The system according to claim 4, characterized in that, The first network module includes: The first network submodule is used to establish a multi-frequency backup link pool and configure a real-time link switching mechanism to complete the networking operation if the probability that the relative displacement speed between the portable base station and the neighboring base station in the probability distribution belongs to the first displacement speed range is greater than the probability threshold. The second network submodule is used to adopt an adaptive link optimization strategy to adjust the antenna beam coverage or transmission power and retain basic backup links if the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution is greater than the probability threshold. The third network submodule is used to establish a fixed high-gain directional link, concentrate beam energy, and retain the minimum redundant link to complete the networking operation if the probability that the relative displacement velocity between the portable base station and the neighboring base station in the probability distribution belongs to the third displacement velocity range is greater than the probability threshold.

Images