Processing method and apparatus for wireless network coverage vulnerability, and computing device

Abstract

Provided in the present application are a processing method and apparatus for a wireless network coverage vulnerability, and a computing device. The method comprises: acquiring weak fingerprint data of a terminal within a coverage range, which data is collected by a wireless access point (AP) in a preset time period, wherein the weak fingerprint data of the terminal is signal data of the AP, which is collected by the terminal in a weak-signal situation; on the basis of the weak fingerprint data of each terminal, determining a suspected coverage vulnerability; on the basis of the suspected coverage vulnerability, determining whether the suspected coverage vulnerability meets a filtering condition; and when the suspected coverage vulnerability meets the filtering condition, determining that the suspected coverage vulnerability is not a wireless network coverage vulnerability.

Description

Wireless network coverage vulnerability processing method and device, computing equipment

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to the Chinese patent application filed with the China Patent Office on September 25, 2023, with application number 202311243121.4 and application name “A method, device and computing device for processing wireless network coverage vulnerabilities”, the entire contents of which are incorporated by reference into this application. Technical Field

[0003] The present application relates to the field of communication technology, and in particular to a method and apparatus for processing wireless network coverage vulnerabilities, a computing device, and a storage medium. Background Art

[0004] In modern society, people are increasingly dependent on wireless networks. The stability and transmission speed of wireless networks directly impact their quality of life and work efficiency. However, wireless network coverage holes still exist for various reasons. A wireless network coverage hole refers to insufficient wireless network signal strength within a specific area, resulting in the inability to provide normal wireless network services.

[0005] Summary of the Invention

[0006] The present invention provides a method for addressing wireless network coverage vulnerabilities.

[0007] In a first aspect, an embodiment of the present application provides a method for processing a wireless network coverage vulnerability, including: obtaining weak fingerprint data of a terminal within a coverage range collected by a wireless access point (AP); the weak fingerprint data of the terminal is signal data of the AP collected by the terminal under weak signal conditions; determining a suspected coverage vulnerability based on the weak fingerprint data of the terminal; based on the suspected coverage vulnerability, determining whether the suspected coverage vulnerability meets a filtering condition; when the suspected coverage vulnerability meets the filtering condition, determining that the suspected coverage vulnerability does not belong to a wireless network coverage vulnerability.

[0008] In the above implementation, signal data collected from each AP within the coverage area under weak signal conditions, i.e., weak fingerprint data for each terminal, is obtained. Based on this weak fingerprint data, suspected coverage holes are identified and judged to determine whether they meet the filtering criteria, thereby eliminating misjudgments of wireless network coverage holes caused by these filtering criteria. Compared to current implementations, this approach considers suspected coverage holes in terms of both time and space, taking into account factors that can easily lead to misjudgment, effectively ensuring the comprehensiveness and accuracy of coverage hole detection.

[0009] In a possible implementation, before obtaining weak fingerprint data of terminals within a coverage area collected by a wireless access point AP within a preset time period, the method further includes:

[0010] When the uplink signal strength of the first terminal is less than a preset threshold, a first AP is triggered to collect weak fingerprint data of the first terminal, where the first AP is a device of a network to which the first terminal accesses.

[0011] In a possible implementation, the filtering condition is used to characterize that the suspected coverage vulnerability is not a wireless network coverage vulnerability from a time dimension and / or a space dimension.

[0012] In one possible implementation, the method further includes: obtaining weak fingerprint data of the terminal within the coverage range collected by a wireless access point (AP); the weak fingerprint data of any terminal is signal data of the AP collected by the terminal under weak signal conditions; determining a suspected coverage vulnerability based on the weak fingerprint data of the terminal and a problematic AP; the problematic AP is an AP whose signal strength is lower than a threshold among the APs associated with each terminal, obtained through the cloud;

[0013] For each suspected coverage vulnerability, it is determined whether the suspected coverage vulnerability meets the filtering condition. If the suspected coverage vulnerability meets the filtering condition, it is determined that the suspected coverage vulnerability does not belong to the wireless network coverage vulnerability.

[0014] In the above implementation method, the signal data of each AP collected under weak signal conditions within the coverage range, that is, the weak fingerprint data of each terminal, is obtained; the suspected coverage vulnerability is determined based on the weak fingerprint data and the AP obtained from the cloud, and then it is judged whether the suspected coverage vulnerability meets the filtering conditions, so as to determine that the suspected coverage vulnerability does not belong to the wireless network coverage vulnerability. In other words, when determining the suspected coverage vulnerability, the suspected coverage vulnerability is determined based on the weak fingerprint data and the connected problematic AP, which is more accurate.

[0015] In one possible implementation, the filtering condition includes that the collection times corresponding to the first coverage hole in at least two collection cycles belong to the same time range, and / or that the AP corresponding to the first coverage hole is an edge AP, wherein the first coverage hole is any suspected coverage hole.

[0016] In the above implementation, the filtering conditions include two aspects. One is the temporal influencing factor, that is, judging the tidal phenomenon, that is, judging whether the corresponding collection moments in at least two collection cycles belong to the same time range, that is, coverage holes that appear at the same time range are tidal phenomena, which is one of the factors that may cause misjudgment of wireless network holes; the other is the spatial influencing factor, that is, judging whether the corresponding AP of the coverage hole belongs to an edge AP. If so, the suspected coverage hole does not belong to a wireless network coverage hole. That is, if the suspected coverage hole meets the filtering conditions, it does not belong to a wireless network coverage hole. Taking the causes that may affect the misjudgment into consideration, the wireless network coverage hole can be effectively handled and its accuracy can be improved.

[0017] In one possible implementation, determining whether a suspected coverage vulnerability meets a filtering condition includes: determining the suspected coverage vulnerability corresponding to different times based on weak fingerprint data of each terminal within the coverage range of each AP obtained at different times; if the difference between the collection times corresponding to any suspected coverage vulnerability in at least two collection cycles is less than a preset time difference threshold, then the suspected coverage vulnerability meets the filtering condition.

[0018] In the above implementation method, the weak fingerprint data of each terminal within the coverage of each AP at different times is obtained, and the suspected coverage holes corresponding to different time periods are determined based on the obtained weak fingerprint data of each terminal. That is, if the difference between at least two acquisition moments corresponding to the suspected coverage hole is greater than the preset time difference threshold, then the suspected coverage hole meets the filtering conditions and is a normal tidal phenomenon. That is to say, the terminal is in a weak signal state due to similar reasons, such as a large number of users using the network during the morning rush hour, causing suspected coverage holes, thereby avoiding misjudgment and improving the accuracy of coverage hole processing.

[0019] In a possible implementation, whether the suspected coverage hole meets the filtering condition includes: if it is determined that the first AP corresponding to the suspected coverage hole is an edge AP, then the suspected coverage hole meets the filtering condition.

[0020] According to the above implementation, when the first AP corresponding to the suspected coverage vulnerability is an edge AP, the filtering condition is met. That is, the suspected coverage vulnerability occurs because the connected AP is an edge AP, which causes a misjudgment, thereby improving the accuracy of wireless network vulnerability determination.

[0021] In one possible implementation, if the collection moments corresponding to the suspected coverage vulnerability in at least two collection cycles do not belong to the same time range, determine whether the first AP corresponding to the suspected coverage vulnerability is an edge AP; if the first AP corresponding to the suspected coverage vulnerability is an edge AP, determine that the suspected coverage vulnerability meets the filtering conditions; or, if the first AP corresponding to the suspected coverage vulnerability is not an edge AP, determine whether the collection moments corresponding to the suspected coverage vulnerability in at least two collection cycles belong to the same time range; if the collection moments corresponding to the suspected coverage vulnerability in at least two collection cycles belong to the same time range, determine that the suspected coverage vulnerability meets the filtering conditions.

[0022] According to the above implementation, a suspected coverage hole meets the filtering criteria if the corresponding collection times in at least two collection cycles fall within the same time range, or if the problematic AP corresponding to the suspected coverage hole is an edge AP. In other words, as long as either of these criteria is met, the filtering criteria is met. Conversely, if the corresponding collection times in at least two collection cycles do not fall within the same time range, the suspected coverage hole is further determined to determine whether the AP corresponding to the suspected coverage hole is an edge AP. If not, the suspected coverage hole does not meet the filtering criteria, meaning that the suspected coverage hole is neither an edge AP nor falls within the same time range.

[0023] In one possible implementation, determining whether the AP corresponding to the suspected coverage loophole meets the filtering conditions includes: determining the node characteristics of each of the first AP and the second AP corresponding to the suspected coverage loophole; the node characteristics of any AP are used to represent the wireless network status of the AP; the second AP includes at least one other AP other than the first AP; determining the inter-node characteristics between the first AP corresponding to the suspected coverage loophole and the second AP; wherein the inter-node characteristics are used to represent the correlation between AP nodes based on terminal roaming data; and determining that the first AP is an edge AP based on the node characteristics of each of the first AP and the second AP and the inter-node characteristics.

[0024] In one possible implementation, determining that the first AP is an edge AP based on the node characteristics of each of the first AP and the second AP and the inter-node characteristics includes: constructing AP graph structure data based on the node characteristics of each of the first AP and the second AP and the inter-node characteristics; and determining that the first AP is an edge AP through the AP graph structure data.

[0025] In the above implementation, when determining whether it is an edge AP, it is mainly determined based on the node characteristics of the corresponding AP and the node characteristics between other APs. First, the node characteristics of each AP used to represent the wireless network status of the AP are determined. Then, based on the correlation between AP nodes based on terminal roaming data, the node characteristics between the corresponding AP and other APs are determined. The node characteristics of each AP and the node characteristics between other APs are used to construct AP graph structure data. Based on the AP graph structure data, the edge AP among each AP is determined. In other words, the above method eliminates the phenomenon that affects the judgment of coverage holes in the spatial dimension, thereby improving the accuracy of coverage hole processing.

[0026] In a possible implementation, determining, according to the node feature of the first AP and the inter-node feature, that the AP corresponding to the suspected coverage hole is an edge AP includes:

[0027] The attention coefficient between the first AP and the second AP is determined based on the node characteristics between the first AP and the second AP, the node characteristics of the first AP, and the node characteristics of the second AP. The fusion node characteristics of the first AP are determined based on the attention coefficients of each adjacent AP. The first AP is determined to be an edge AP based on the fusion node characteristics.

[0028] In the above implementation, for the AP graph structure, the attention coefficient between each AP and other APs is determined based on the node features of each AP and the inter-node features between nodes, which helps capture the interactive relationships between AP nodes. After obtaining the attention coefficient, the AP's fused node features are determined based on the attention coefficient. This means that the original AP node features are updated to the new features of the AP node, namely the AP's fused node features. Finally, based on the AP's fused node features, the AP corresponding to the suspected coverage hole is determined to be an edge AP. This avoids the weak signal conditions introduced by edge APs, which can cause misjudgment of coverage holes, and improves the accuracy of coverage hole processing.

[0029] In a possible implementation, edge APs among APs are determined using AP graph structure data, including: inputting the AP graph structure data into a graph structure model to obtain a node feature vector of the AP; and inputting the node feature vector of the AP into a classification model to obtain that the first AP is an edge AP.

[0030] In the above implementation, edge APs are identified based on the AP graph data using a graph model and a classification model. Specifically, the AP graph data is first input into the graph model to obtain node feature vectors for each AP. These vectors are then input into the classification model to identify APs suspected of having coverage holes as edge APs. Because the model's parameters are trained, edge APs are more accurately identified.

[0031] In one possible implementation, the node characteristics include one or more of AP mutual scanning data, roaming data of each terminal under AP coverage, and online and offline data; the inter-node characteristics are the number of roaming times of terminals between nodes and / or the roaming direction of the terminals.

[0032] In the above implementation, the AP's node characteristics are determined based on AP inter-scanning data, roaming data, and online and offline data of each terminal within the AP's coverage area. The AP's inter-node characteristics are determined based on the number of inter-node roaming times and / or roaming directions of terminals. Determining node characteristics and inter-node characteristics using this data more effectively derives the true relationship between AP characteristics, resulting in more accurate calculations for determining coverage gaps.

[0033] In one possible implementation, when the coverage range of the AP includes multiple terminals, suspected coverage loopholes are determined based on the weak fingerprint data of the multiple terminals, including: clustering the weak fingerprint data of the multiple terminals based on a density parameter to obtain one or more first clusters and at least one first outlier; the first outlier is weak fingerprint data that does not belong to any first cluster; based on the adjusted density parameter, clustering the multiple first clusters and the first outlier to obtain one or more second clusters and at least one second outlier until a clustering termination condition is met; the second outlier is weak fingerprint data that does not belong to any second cluster; each cluster that meets the clustering termination condition is determined as a suspected coverage loophole; wherein the density parameter includes the cluster domain size and the minimum number of points in the cluster; the clustering process is based on the cosine distance of the weak fingerprint data; the adjusted density parameter is the cluster domain size and the minimum number of points in the cluster redefined based on the multiple first clusters and the multiple first outliers.

[0034] In the above implementation, the weak fingerprint data of each terminal is clustered using a density parameter to determine suspected coverage loopholes. That is, the weak fingerprint data of each terminal is clustered based on the density parameter determined by the cluster domain size and the minimum number of points in the cluster to obtain each first cluster and each first outlier. Then, the clustering process is performed again based on the adjusted density parameter to determine whether each second cluster and each second outlier after clustering meets the clustering termination condition. If not, the clustering process is continued. If so, each cluster that meets the clustering termination condition is determined as a suspected coverage loophole. This can achieve efficient and accurate processing of coverage loophole areas with high accuracy and low false alarm rate. Among them, the clustering process is based on the cosine distance of the weak fingerprint data, making it more robust.

[0035] In a possible implementation, the clustering termination condition is determined based on the comparison of outliers in at least two clustering results. If the number of outliers is equal, the clustering termination condition is met; otherwise, clustering continues.

[0036] In the above implementation method, the abnormal points in the results of at least two clustering calculations are used to determine whether the clustering is terminated, that is, the termination conditions are determined, and the final cluster is determined according to the termination conditions, thereby obtaining suspected coverage loopholes, improving the accuracy of determining suspected coverage loopholes, and reducing the probability of misjudgment.

[0037] In a possible implementation, after determining that the suspected coverage loophole meets the filtering condition, the method further includes: determining whether the number of terminals meets a threshold; if so, finally determining that the suspected coverage loophole is a wireless network coverage loophole.

[0038] In the above implementation method, before determining whether it is a wireless network coverage vulnerability, it is also necessary to confirm the number of terminals in the suspected coverage vulnerability at this time. If the number meets the threshold, it is a wireless network coverage vulnerability. This can effectively avoid the misjudgment of wireless network coverage vulnerabilities caused by the terminal itself, and ensure the accuracy of the determination of wireless network coverage vulnerabilities.

[0039] In a possible implementation, the weak signal condition is that the signal strength of the AP perceived by the terminal is less than a preset threshold.

[0040] In a possible implementation, after determining whether the suspected coverage vulnerability meets the filtering condition, the method further includes:

[0041] When the suspected coverage vulnerability does not meet the filtering condition, it is determined that the suspected coverage vulnerability belongs to a wireless network coverage vulnerability.

[0042] In a second aspect, an embodiment of the present application provides a device for handling a wireless network coverage vulnerability, including:

[0043] An acquisition unit is configured to acquire weak fingerprint data of a terminal within a coverage range collected by a wireless access point AP; the weak fingerprint data of the terminal is signal data of the AP collected by the terminal under weak signal conditions;

[0044] a determining unit, configured to determine a suspected coverage vulnerability based on the weak fingerprint data of the terminal;

[0045] The determining unit is further configured to determine, based on the suspected coverage vulnerability, whether the suspected coverage vulnerability meets a filtering condition; if the suspected coverage vulnerability meets the filtering condition, determine that the suspected coverage vulnerability does not belong to a wireless network coverage vulnerability.

[0046] In one possible implementation, the acquisition unit is further configured to acquire weak fingerprint data of the terminal within the coverage range collected by a wireless access point (AP); the weak fingerprint data of any terminal is signal data of the AP collected by the terminal under weak signal conditions; a suspected coverage vulnerability is determined based on the weak fingerprint data of the terminal and a problematic AP; the problematic AP is an AP associated with the terminal obtained from the cloud whose signal strength is lower than a threshold;

[0047] The determining unit is further configured to determine, for each suspected coverage vulnerability, whether the suspected coverage vulnerability meets a filtering condition, and if the suspected coverage vulnerability meets the filtering condition, determine that the suspected coverage vulnerability does not belong to a wireless network coverage vulnerability.

[0048] In one possible implementation, the filtering condition includes that the collection times corresponding to the first coverage hole in at least two collection cycles belong to the same time range, and / or that the AP corresponding to the first coverage hole is an edge AP, wherein the first coverage hole is any suspected coverage hole.

[0049] In a possible implementation, the determining unit is configured to determine suspected coverage holes corresponding to different moments based on weak fingerprint data of each terminal within the coverage range of each AP acquired at different moments;

[0050] The determining unit is configured to determine that any suspected coverage hole meets the filtering condition if a difference between the collection times corresponding to the suspected coverage hole in at least two collection cycles is less than a preset time difference threshold.

[0051] In a possible implementation, the determining unit is further configured to determine that, if the first AP corresponding to the suspected coverage hole is an edge AP, the suspected coverage hole meets a filtering condition.

[0052] In one possible implementation, the determination unit is used to determine whether the first AP is an edge AP if the collection moments corresponding to the suspected coverage vulnerability in at least two collection cycles do not belong to the same time range; if the first AP is an edge AP, determine that the suspected coverage vulnerability meets the filtering conditions; or, if the first AP is not an edge AP, determine whether the collection moments corresponding to the suspected coverage vulnerability in at least two collection cycles belong to the same time range; if the collection moments corresponding to the suspected coverage vulnerability in at least two collection cycles belong to the same time range, determine that the suspected coverage vulnerability meets the filtering conditions.

[0053] The determining unit is configured to determine node characteristics of each of the first AP and the second AP; the node characteristics of any AP are used to represent the wireless network status of the AP; the second AP includes at least one other AP except the first AP;

[0054] The determining unit is configured to determine an inter-node feature between the first AP and the second AP corresponding to the suspected coverage hole; wherein the inter-node feature is used to represent the correlation between AP nodes based on terminal roaming data;

[0055] The determining unit is configured to determine that the first AP is an edge AP according to respective node characteristics of the first AP and the second AP and the inter-node characteristics.

[0056] In a possible implementation, the determining unit is configured to construct AP graph structure data according to respective node characteristics of the first AP and the second AP and the inter-node characteristics; and determine that the first AP is an edge AP through the AP graph structure data.

[0057] In one possible implementation, the determining unit is configured to determine an attention coefficient between the first AP and the second AP based on the node feature of the first AP and the inter-node feature of the second AP, and the node feature of the first AP and the node feature of the second AP; and determine a fused node feature of the first AP based on the attention coefficients of each adjacent AP;

[0058] The determining unit is configured to determine, based on the fusion node feature, that the first AP is an edge AP.

[0059] In a possible implementation, the determining unit is further configured to input the AP graph structure data into a graph structure model to obtain a node feature vector of the AP;

[0060] The determining unit is further configured to input the node feature vector of the AP into a classification model to determine that the first AP is an edge AP.

[0061] In a possible implementation manner, the node characteristics include one or more of AP mutual scanning data, roaming data of each terminal under AP coverage, and online and offline data;

[0062] The inter-node feature is the number of roaming times of the terminal between the nodes and / or the roaming direction of the terminal.

[0063] In a possible implementation, when the coverage area of ​​the AP includes multiple terminals, the determining unit is configured to perform clustering processing on the weak fingerprint data of the multiple terminals based on a density parameter to obtain one or more first clusters and at least one first outlier; the first outlier is weak fingerprint data that does not belong to any first cluster;

[0064] The determining unit is configured to cluster the plurality of first clusters and the first outlier based on the adjusted density parameter to obtain one or more second clusters and at least one second outlier until a clustering termination condition is satisfied; the second outlier is weak fingerprint data that does not belong to any second cluster;

[0065] The determination unit is configured to determine each cluster as a suspected coverage hole when a cluster termination condition is satisfied; wherein the density parameters include a cluster domain size and a minimum number of points in a cluster; the clustering process is based on a cosine distance of weak fingerprint data; and the adjusted density parameters are the cluster domain size and the minimum number of points in a cluster redefined based on the plurality of first clusters and the plurality of first outliers.

[0066] In a possible implementation, the determination unit determines the clustering termination condition based on a comparison of outliers in at least two clustering results. If the number of outliers is equal, the clustering termination condition is met; otherwise, clustering continues.

[0067] In a possible implementation, the determining unit is further configured to determine whether the number of terminals meets a threshold value, and if so, ultimately determine that the suspected coverage loophole is a wireless network coverage loophole.

[0068] In a third aspect, the present application further provides a computing device, comprising:

[0069] a memory for storing program instructions;

[0070] The processor is configured to call program instructions stored in the memory and execute the method for handling the wireless network coverage vulnerability according to the obtained program.

[0071] In a fourth aspect, the present application also provides a computer-readable non-volatile storage medium, comprising computer-readable instructions. When a computer reads and executes the computer-readable instructions, the computer executes the above-mentioned method for processing wireless network coverage vulnerabilities.

[0072] The technical effects that can be achieved in any of the second to fourth aspects mentioned above can refer to the description of the beneficial effects in the first aspect mentioned above, and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0074] FIG1 is a framework diagram of a system for processing wireless network coverage vulnerabilities provided in an embodiment of the present application.

[0075] FIG2 is a flow chart of a method for handling wireless network coverage vulnerabilities provided in an embodiment of the present application.

[0076] FIG3 is a flowchart of a trigger for obtaining weak fingerprint data provided by an embodiment of the present application.

[0077] FIG4 is a flow chart of a method for determining a suspected coverage vulnerability provided by an embodiment of the present application.

[0078] FIG5 is a schematic diagram of an AP graph structure provided in an embodiment of the present application.

[0079] FIG6 is a schematic diagram of determining edge APs among APs according to an embodiment of the present application.

[0080] FIG7 is a schematic diagram of the overall process of determining coverage vulnerabilities in an embodiment of the present application.

[0081] FIG8 is a schematic diagram of a device for processing wireless network coverage vulnerabilities provided in an embodiment of the present application.

[0082] FIG9 also provides a computing device according to an embodiment of the present application. DETAILED DESCRIPTION

[0083] .

[0084] To make the purpose, technical solutions, and advantages of the embodiments of this application more clear, the technical solutions of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of the technical solutions of this application, but not all of them. Based on the embodiments described in this application document, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the technical solutions of this application.

[0085] The terms "first" and "second" in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the term "comprising" and any of its variations are intended to cover non-exclusive protection. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices. "Multiple" in this application can mean at least two, for example, two, three or more, and the embodiments of this application are not limited thereto.

[0086] In addition, the term "and / or" in this document simply describes an association between related objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document, unless otherwise specified, generally indicates an "or" relationship between the related objects.

[0087] With the development of society, wireless networks have become an indispensable part of people's daily lives and work. Whether it's online learning, remote work, social media, or online entertainment, wireless networks are essential. However, due to various reasons, wireless network coverage holes still exist in many areas. This not only affects people's online experience, but can also impact important network services such as telemedicine and online education. A wireless network coverage hole refers to a situation where the wireless network signal strength in a specific area is insufficient to ensure normal communication, resulting in an inability to use wireless network services. This can lead to slow network connection speeds, unstable data transmission, or even complete network connection failure. Current approaches often use access points (APs) to obtain the signal strength of clients (i.e., terminals). The signal strengths at each AP location and client are compared with a preset signal strength threshold to identify suspected coverage holes. The number of clients at the location corresponding to the suspected coverage hole is then determined, and the coverage hole is determined based on the results. However, this coverage hole handling process does not account for cases where coverage holes are misidentified, such as tidal waves or edge APs. Tidal waves refer to the increased demand for wireless network access during rush hour, resulting in slower network speeds. For example, during rush hour, users' phones may continue to connect to their work or home APs even after leaving their workplaces, resulting in persistently low signal levels. These weak signal events aren't caused by coverage holes, but rather by user behavior. Therefore, a lack of analysis of the tidal effect can lead to misjudgments of coverage holes.

[0088] Edge APs are wireless access points located at the edge of network coverage. They are typically located at the outermost edge of the network or farther from core APs, such as near building walls, entrances, or parking lots. Even when users have left their work or living areas but are near building walls, such as parking lots, entrances, or home fronts, they may still be connected to edge APs. This can introduce weak signal events and lead to false detection of coverage holes.

[0089] The processing of coverage holes is bound to have inaccuracies. The embodiment of the present application provides a method for processing wireless network coverage holes, which is used to improve the accuracy of processing coverage holes.

[0090] Figure 1 is a framework diagram of a wireless network coverage vulnerability processing system provided in an embodiment of the present application, and the system includes a device AP101 and a terminal 102. The device AP101 is provided with an AP, i.e., an AP node, which is used to provide a network for the terminal 102 to achieve interaction. The terminal 102 is used to implement network services after accessing the AP node. The terminal can be a mobile phone, smart wearable device, tablet computer, notebook or other terminal device. Figure 1 only shows one device AP as an example, and the number is not limited. The device AP101 and the terminal 102 use wireless connection, and other connection methods may also be used without specific limitation here.

[0091] The present application also discloses a method for processing wireless network coverage vulnerabilities, as shown in FIG2 . FIG2 is a flow chart of a method for processing wireless network coverage vulnerabilities provided by the present application, including:

[0092] Step 201: Acquire weak fingerprint data of each terminal within the coverage area collected by each wireless access point AP; the weak fingerprint data of any terminal is the signal data of each AP collected by the terminal under weak signal conditions.

[0093] To ensure that there are no coverage holes within the wireless network, coverage holes need to be addressed. However, addressing coverage holes requires determining the signal strength of the APs within the wireless network's coverage area. Therefore, it is necessary to obtain weak fingerprint data from each terminal within the coverage area, collected by each wireless access point (AP). This data is the signal data collected by the terminal from each AP under weak signal conditions. A weak signal condition refers to a situation where the signal strength of each AP perceived by the terminal is less than a threshold, in which case the terminal is considered to be in a weak signal condition. The threshold can be an empirical value set based on actual needs. For example, if the perceived signal strength of a particular AP is less than -40dBm, the terminal is considered to be in a weak signal condition.

[0094] As shown in Figure 3, Figure 3 provides a flowchart for triggering the acquisition of weak fingerprint data in an embodiment of the present application. Specifically, when a wireless access point (AP) discovers that a terminal's uplink signal is less than a threshold set based on actual needs, it sends a request signal to the terminal based on the 802.1 protocol. After receiving the request signal, the terminal responds to the wireless access point (AP) with the downlink signal of each AP on the terminal side in the form of a report. During this process, due to signal fluctuations, multiple request signals are sent and replies are received in the form of reports to obtain the weak fingerprint data of each terminal.

[0095] In one example, assume there are three APs in this scenario. Of course, the network environment provided by this application is not limited to a number of wireless access points (APs), and can be multiple. In this embodiment, three APs, namely AP1, AP2, and AP3, are used as an example. At time t1, it is detected that the uplink signal of the terminal is less than the threshold, triggering the 802.11K protocol to send a request signal. The terminal returns the 802.11K protocol report, and the weak signal data at t1 is obtained: AP1: -78; AP2: -80; AP3: -76; where the values ​​(-78, -80, -76) represent the signal strength. Similarly, at different times t2 and t3, the interaction is still carried out, and the weak signal data at t2 is obtained: AP1: -82; AP2: -; AP3: -80; and the weak signal data at t3 is obtained: AP1: -90; AP2: -; AP3: -83; and the final weak fingerprint data is obtained by superimposing them in chronological order: AP1: -83; AP2: -80; AP3: -79.

[0096] It should be noted that the above-mentioned superposition adopts the method of adding and averaging, and other methods of superposition can also be used, such as weighted summing and averaging, which is not the only limitation here; the above-mentioned order adopts the time order, and other orders can also be used, which is not the only limitation here.

[0097] S202. Determine suspected coverage vulnerabilities based on the weak fingerprint data of each terminal.

[0098] After determining the weak fingerprint data of each terminal in step S201, a suspected coverage vulnerability is determined based on the weak fingerprint data of each terminal. A suspected coverage vulnerability is a preliminary determination of the possibility of a coverage vulnerability.

[0099] In another embodiment, as shown in FIG4 , FIG4 is a flow chart of a method for determining a suspected coverage vulnerability according to an embodiment of the present application, including:

[0100] S401. Clustering the weak fingerprint data of each terminal based on a density parameter to obtain first clusters and first outliers; each first outlier is weak fingerprint data that does not belong to any first cluster.

[0101] Clustering refers to the use of morphological operators to cluster and merge adjacent, similarly classified regions. A cluster is the name of the region being clustered and merged. In this embodiment of the present application, fingerprint data from each terminal is clustered based on a density parameter to obtain first clusters and first outliers in weak fingerprint data that does not belong to any first cluster.

[0102] In an example, assuming there are 100 weak fingerprint data, these 100 weak fingerprint data are clustered to obtain 15 clusters and 25 first outliers.

[0103] S402. Based on the adjusted density parameter, cluster the first clusters and the first outliers to obtain the second clusters and the second outliers, until a clustering termination condition is met.

[0104] After obtaining the first clusters and first outliers from the initial clustering process, clustering needs to be performed again. This is because the initial clustering process requires a termination condition to be determined, and this condition requires comparison of at least two clustering results. Therefore, after obtaining the first clusters and first outliers, clustering is performed again based on the adjustment of the density parameters to obtain the second clusters and second outliers. The second outliers are then compared with the first outliers. If no changes have occurred, the termination condition is met. If not, clustering continues until the termination condition is met.

[0105] Continuing with the above example, the 15 clusters obtained from the first clustering process are averaged to obtain the first average cluster. The first average cluster and the first outlier are clustered based on the adjusted density parameter to obtain 13 clusters and 25 second outliers after clustering. At this time, the first outlier is equal to the second outlier, and the clustering termination condition is met.

[0106] S403. Determine each cluster that meets the clustering termination condition as a suspected coverage hole; wherein the density parameters include the cluster domain size and the minimum number of points in the cluster; the clustering process is based on the cosine distance of the weak fingerprint data.

[0107] In another embodiment, each cluster that meets the clustering termination condition in step S402 is identified as a suspected coverage hole. The density parameter is determined based on the cluster domain size ε and the minimum number of points in the cluster, MinPts. The initial value is determined based on actual scenarios. Specifically, the cluster domain size ε is empirically set based on the distribution of weak fingerprint data on the time axis, while the minimum number of points in the cluster, MinPts, is also empirically set. The adjusted density parameter is then re-set based on the density parameter using empirical values ​​based on the distribution of each first cluster and each first outlier on the time axis. This can also be understood as an optimization process. In other words, each cluster and each outlier is continuously updated to serve as the initial value for the density parameter. The setting of the acquired weak fingerprint data based on the cluster domain size ε and the minimum number of points in the cluster, MinPts, is based on actual needs. The principle of clustering is to use the cluster domain size ε as the demarcation range. A cluster is formed when the demarcation range reaches the minimum number of points in the cluster, MinPts. It should be noted that the cluster size ε is determined by the cosine value, that is, the smaller the angle, the more similar the features are, and the stronger the correlation of the weak fingerprint data.

[0108] In one example, ε is set to 0.2 based on the cosine distance (cosine value) and MinPts is set to 3 to determine the cosine distance value of each weak fingerprint data. This means that if there are fingerprint data with a cosine distance less than or equal to the set threshold, and the minimum number of points in the cluster involved, MinPts, is greater than 3, then these fingerprint data with similar cosine distances can be clustered together. It should be noted that the initial value of 0.2 for the cosine distance is an empirical value.

[0109] The embodiment of the present application also provides another method for determining the suspected coverage vulnerability in step S202, that is, determining the suspected coverage vulnerability based on weak fingerprint data and problem AP; the problem AP is an AP whose signal strength is lower than a threshold value among the APs associated with each terminal obtained through the cloud. The method for determining the problem AP is that the cloud detects the network experience status of each terminal, and obtains weak fingerprint data and problem AP based on the network experience status, that is, the signal strength of the corresponding AP connected to the terminal is poor, that is, the AP whose signal strength is lower than the threshold value among the APs, and determines the suspected coverage vulnerability based on the weak fingerprint data and problem AP. In short, after the suspected coverage vulnerability is determined, it is possible to know which corresponding AP is connected at this time, so as to prepare for the subsequent judgment of whether it is a wireless network vulnerability.

[0110] S203. For each suspected coverage vulnerability, determine whether the suspected coverage vulnerability meets the filtering condition; if the suspected coverage vulnerability meets the filtering condition, determine that the suspected coverage vulnerability does not belong to the wireless network coverage vulnerability.

[0111] When a suspected coverage vulnerability is determined, the suspected vulnerability is judged mainly based on filtering conditions. If the suspected coverage vulnerability meets the filtering conditions, it is determined that the suspected coverage vulnerability does not belong to a wireless network coverage vulnerability. The filtering conditions are further screened based on the time dimension and the spatial dimension. First, from the time dimension, it is determined whether the suspected coverage vulnerability is a tidal phenomenon; second, from the spatial dimension, it is determined whether the corresponding AP of the suspected coverage vulnerability is an edge AP. If the suspected coverage vulnerability is a tidal phenomenon, that is, the corresponding collection moments in at least two collection cycles belong to the same time range, and / or the corresponding AP of the suspected coverage vulnerability is an edge AP, that is, the filtering conditions are determined based on the above two factors, then the suspected coverage vulnerability obtained at this time is a wireless network coverage vulnerability, effectively improving the accuracy of coverage vulnerability processing.

[0112] In one example, there is a suspected coverage vulnerability M. The collection times are 12:00 and 13:00 respectively. At this time, 12:00 and 13:00 both belong to the time range of 11:00-13:00, that is, they belong to the same time range. It can be determined that it meets one of the filtering conditions.

[0113] In addition to the aforementioned considerations of misjudgments in the temporal and spatial dimensions, the number of terminals must also be taken into account to effectively avoid misjudgments caused by a single terminal experiencing a problem—in other words, misjudgments caused by the terminal itself rather than the network. Furthermore, the number of terminals must be determined to meet a threshold. If so, the suspected coverage vulnerability is ultimately determined to be a wireless network coverage vulnerability. The threshold is set based on empirical values.

[0114] In an example, assume that there are suspected coverage holes A, B, and C. After judging suspected coverage holes A, B, and C, it is found that suspected coverage holes A and B are neither always present suspected coverage holes nor belong to edge APs. Therefore, suspected coverage holes A and B are wireless network coverage holes.

[0115] Among them, there are various implementation methods for determining whether a suspected coverage vulnerability meets the filtering conditions. The first implementation method is provided below, that is, after determining the suspected coverage vulnerability, it is determined whether it is a tidal phenomenon. If so, it meets the filtering conditions, and there is no need to judge other conditions, that is, based on the weak fingerprint data of each terminal within the coverage range of each AP obtained at different times, the suspected coverage vulnerability corresponding to different times is determined; if the difference between the collection times corresponding to any suspected coverage vulnerability in at least two collection cycles is less than the preset time difference threshold, then the suspected coverage vulnerability meets the filtering conditions.

[0116] The preset time difference threshold is set according to the actual experience value, that is, the loopholes that appear within the preset time difference threshold are suspected loopholes caused by tidal phenomena. When defining the preset time difference threshold in one embodiment, it is generally in hours. Of course, other units can also be set, and there is no unique limitation here. For example, that is to say, if the difference between at least two acquisition moments corresponding to the first coverage loophole is greater than the preset time difference threshold, it does not belong to a tidal phenomenon. The so-called tidal phenomenon refers to the suspected coverage loophole obtained by subsequent calculation of weak fingerprint data that appears due to the increase in user demand for using wireless networks, resulting in a slow network speed, such as during rush hour. That is to say, the weak fingerprint data contains time features, so the time distribution of the weak fingerprint data can be known, thereby judging whether the coverage loophole is caused by a tidal phenomenon.

[0117] In one example, the preset time period is set to one week, the collection cycle is one day, and the time of day is set from 0:00 to 24:00. Assuming the preset time difference threshold is defined as 1 hour, if the collection times of a suspected coverage vulnerability are 8:00 on Monday and 8:00 on Tuesday, respectively, and the difference between the two collection times is 8-8 = 0, which is less than the preset time difference threshold, then it is determined to be a tidal phenomenon, meaning that the suspected coverage vulnerability meets the filtering criteria, but the suspected coverage vulnerability is not a wireless network coverage vulnerability. Similarly, if the two collection times of the suspected coverage vulnerability are 9:00 on Monday and 12:00 on Tuesday, respectively, and the difference between the two collection times is 12-9 = 3, which is greater than the preset time difference threshold, then it is not a tidal phenomenon, meaning that the suspected coverage vulnerability is a wireless network coverage vulnerability. If the two collection times of the suspected coverage vulnerability are 11:00 on Monday and 12:00 on Tuesday, respectively, the difference between the two collection times is 12-11 = 1, which is equal to the preset time difference threshold, then it is not a tidal phenomenon and does not meet the filtering criteria for suspected coverage vulnerabilities.

[0118] It should be noted that the above example is based on two examples of the time when suspected coverage holes are collected. Three or more times can also be selected to reduce errors. They are not listed one by one here.

[0119] In one embodiment, a single filtering condition is used to determine whether the difference between at least two acquisition times corresponding to the first coverage hole is equal to a preset time difference threshold. Alternatively, the determination of whether the AP corresponding to the suspected coverage hole is an edge AP may be performed. A second embodiment is provided below, in which if the AP corresponding to the suspected coverage hole is determined to be an edge AP, the suspected coverage hole meets the filtering condition. When determining whether the AP corresponding to the suspected coverage hole is an edge AP as the only filtering condition, that is, after determining the suspected coverage hole, the determination of whether the AP corresponding to the suspected coverage hole is an edge AP is performed. If the AP is an edge AP, the filtering condition is met, and no further determination of other conditions is performed.

[0120] That is, determine whether the corresponding AP of the coverage hole is an edge AP, where the corresponding AP is a suspected coverage hole determined based on weak fingerprint data, or a suspected coverage hole determined based on weak fingerprint data and a problem AP, and the method for determining the problem AP is as shown above in this application.

[0121] That is, after identifying a suspected coverage hole, determine whether the AP corresponding to the suspected coverage hole is an edge AP. If so, the filtering condition is met. There are various ways to identify edge APs. One implementation example below includes determining the node characteristics of each AP; the node characteristics of any AP are used to represent the wireless network status of that AP; determining inter-node characteristics between APs; wherein the inter-node characteristics are used to represent the association between AP nodes based on terminal roaming data; constructing AP graph structure data based on the node characteristics of each AP and the inter-node characteristics between APs; and identifying edge APs within each AP using the AP graph structure data.

[0122] Roaming data refers to the relevant data of the terminal roaming between different APs in the wireless network. When determining whether any AP is an edge AP, it is determined based on the node characteristics of the AP and the inter-node characteristics of the AP. The node characteristics of each AP and the inter-node characteristics of the AP are taken into account, which effectively ensures the accuracy of the correlation. In one embodiment, the wireless network status of each AP is first determined, that is, the node characteristics of the AP; and the node characteristics of the AP include AP mutual scanning data, roaming data of each terminal under the coverage of the AP, and online and offline data; after determining the node characteristics of the AP, the correlation of the terminal roaming data between each AP is determined, that is, the inter-node characteristics of the AP; and the inter-node characteristics of the AP are the number of roaming times of the terminal between nodes and / or the roaming direction of the terminal. After determining the node characteristics of the AP and the inter-node characteristics of the AP, the AP graph structure data is constructed, and the edge AP in each AP is determined based on the AP graph structure data. Among them, the node characteristics of the above-mentioned AP are not fixed and can be the data in the above-mentioned embodiment or other data.

[0123] As shown in Figure 5, Figure 5 provides a schematic diagram of an AP graph structure for an embodiment of the present application. Figure 5 shows five APs in a wireless network, as well as edge features between the APs and other APs, that is, roaming data. The larger the edge feature, the more frequent the roaming times between the two APs, and the closer the relationship between the APs. Based on Figure 5, the node features of the AP include AP mutual scanning data. The so-called AP mutual scanning data is the signal strength data scanned by the APs to each other, which is divided into 5G frequency band (5G_rssi) and 2.4G frequency band (2.4G_rssi), and also includes the number of times the AP is scanned (Scaned_times); in addition to AP mutual scanning data, roaming data is also involved, namely the signal strength of the terminal roaming into the AP (Roam_in_rssi) and the signal strength of the terminal roaming out of the AP (Roam_out_rssi); and online and offline data, namely the number of times the AP is the first AP associated with the terminal (First Online_count), the number of times the AP is the last AP associated with the terminal (Last Offline_count), and the number of terminals associated with the AP (Sta_count).

[0124] The determination of edge APs within each AP based on the constructed AP graph structure data is primarily based on the graph structure model and the classification model. In one embodiment, the AP graph structure data is first input into the graph structure model to obtain the node feature vector for each AP. Next, each AP's node feature vector is input into the classification model to determine the edge APs within each AP. The output of the classification model is a two-dimensional vector (X, Y), where X represents the probability that the node belongs to the first category and Y represents the probability that it belongs to the second category. By selecting the larger of these two options, the node type is determined, thereby determining the edge APs within each AP.

[0125] In one example, as shown in FIG6 , FIG6 is a schematic diagram of determining edge APs among APs provided by an embodiment of the present application. AP graph structure data is input into a graph structure model (such as a graph attention model), and calculation and output are performed in the graph attention model, with the output being a node feature vector of each AP. The node feature vector of each AP is input into a classification model to obtain a cross-entropy loss function. Based on the cross-entropy loss function, the parameters of the classification model are trained, i.e., the final output is (X, Y). X is pre-set to represent an edge AP and Y to represent a non-edge AP. Assuming that X is 0.6 and Y is 0.4 at this time, and X>Y, the final output is X, i.e., the AP is an edge AP.

[0126] It should be noted that the graph structure model can be a different type of model and the classification model can also be a different type of model. In the embodiment of this application, a binary classifier is used for example. The embodiment of this application takes the graph attention model as an example, and other models can also be used, such as a tree graph structure model, a multi-view graph structure model, etc., which are not specifically limited here.

[0127] There are also various processing principles based on the internal structure of the model. The following takes the principle of the graph attention model as an example. For any AP in the AP graph structure data, the attention coefficient between the AP and the adjacent AP is determined based on the node features between the AP and the adjacent AP, the node features of the AP, and the node features of the adjacent AP. The fused node features of the AP are determined based on the attention coefficients of each adjacent AP. The edge APs in each AP are determined based on the fused node features of each AP.

[0128] The AP graph structure data includes node features between an AP and its neighbors, the AP's node features, and the node features of its neighbors. These three features are combined to determine the attention coefficient between the AP and its neighbors. The attention coefficient effectively represents the relationship between nodes. After obtaining the attention coefficient, the AP's fused node features are determined based on the attention coefficients of each neighboring AP. This is done by updating the AP's node features to obtain the AP's fused node features. Based on the fused node features of each AP, edge APs are identified within each AP.

[0129] In an example, assuming that there are AP1, AP2, and AP3, when calculating the fused node feature of AP1, the attention coefficient 1 and attention coefficient 2 of AP1 and AP2 are first obtained based on the inter-node features of AP1 and AP2, the inter-node features of AP1 and AP3, and the node features of AP1, AP2, and AP3. The fused node feature m of AP1 is determined based on the attention coefficient 1 and the attention coefficient 2, and whether AP1 is an edge AP is determined based on the fused node feature m.

[0130] In the first embodiment described above, whether the corresponding collection moments of the first coverage hole in at least two collection cycles belong to the same time range is used as a single filtering condition for single-aspect judgment. The second implementation method is to judge whether the corresponding AP of the suspected coverage hole is an edge AP. When both judgments are taken into consideration, in the first embodiment, when the corresponding collection moments of the first coverage hole in at least two collection cycles do not belong to the same time range, that is, the suspected coverage hole does not conform to the tidal phenomenon, it is necessary to continue to judge according to the second judgment condition, that is, to judge whether the corresponding AP of the suspected coverage hole belongs to an edge AP. If it does not belong to an edge AP, it does not meet the filtering condition. If the corresponding collection moments in at least two collection cycles belong to the same time range, that is, it meets the filtering condition, it is also possible to continue to judge whether the suspected coverage hole is an edge AP, thereby increasing the accuracy of its judgment. That is, if the suspected coverage hole neither conforms to the tidal phenomenon nor is an edge AP, it is a wireless coverage hole, making it more accurate to determine.

[0131] It should be noted that after determining the suspected coverage hole, it is possible to simultaneously determine whether the corresponding collection moments in at least two collection cycles belong to the same time range and whether the suspected coverage hole is an edge AP. It is also possible to first determine whether the corresponding collection moments in at least two collection cycles belong to the same time range, and then determine whether they are all edge APs; it is also possible to first determine whether the suspected coverage hole is an edge AP, and then determine whether the corresponding collection moments of the suspected coverage hole in at least two collection cycles belong to the same time range. The order is not uniquely limited.

[0132] As shown in FIG7 , FIG7 is a schematic diagram of the overall process of determining coverage vulnerabilities in an embodiment of the present application. The first implementation method in the present application is as follows:

[0133] The cloud periodically collects relevant wireless network data. Based on this data, weak fingerprint data of each terminal within the coverage area collected by each wireless access point (AP) is obtained. The weak fingerprint data is clustered to obtain suspected coverage vulnerabilities. The suspected vulnerabilities are then determined to meet filtering criteria. If they meet the filtering criteria, the suspected coverage vulnerabilities are determined not to be wireless network coverage vulnerabilities. This is done by determining whether there is a tidal phenomenon, i.e., the corresponding collection times in at least two collection cycles fall within the same time range. If so, it is a tidal phenomenon, and the AP is then determined to be an edge AP. If so, the number of terminals at that time is obtained, and a determination is made as to whether the number of terminals reaches a threshold. If so, the suspected coverage vulnerabilities are determined not to be wireless network coverage vulnerabilities, thus eliminating misjudgments caused by tidal phenomena and edge APs. The determination of the number of terminals considers that if multiple terminals under an AP have suspected coverage vulnerabilities after eliminating tidal phenomena and edge APs, then a wireless network coverage vulnerability exists, thus avoiding misjudgments. For edge APs, AP graph structure data constructed based on the AP's node characteristics and inter-node characteristics is input into a graph structure model and a classification model to determine the AP type, i.e., whether it is an edge AP.

[0134] Another implementation method is also provided in the present application as follows: the cloud regularly collects relevant data of the wireless network, obtains weak fingerprint data of each terminal within the coverage range collected by each wireless access point AP based on the relevant data, determines whether the AP connected to each terminal is a problem AP, determines that the AP connected to each terminal is a problem AP, and obtains suspected coverage vulnerabilities based on the problem AP and the weak fingerprint data, wherein the problem AP is the network usage status of each terminal obtained by the cloud, and obtains an AP whose signal strength is lower than the signal threshold based on the network usage status. The suspected vulnerability is determined to be valid for filtering conditions. If the suspected coverage vulnerability meets the filtering conditions, it is determined that the suspected coverage vulnerability is not a wireless network coverage vulnerability. This is done by determining whether there is a tidal phenomenon, i.e., the corresponding collection times in at least two collection cycles fall within the same time range. If so, it is a tidal phenomenon, and the AP is then determined to be an edge AP. If so, the number of terminals at that time is obtained to determine whether the number of terminals reaches a threshold. If so, the suspected coverage vulnerability is determined to be not a wireless network coverage vulnerability, thus eliminating misjudgments caused by tidal phenomena and edge APs. The determination of the number of terminals is based on the consideration that if multiple terminals under an AP have suspected coverage vulnerabilities after eliminating tidal phenomena and edge APs, then a wireless network coverage vulnerability exists, thus avoiding misjudgments. The determination of edge APs involves inputting AP graph structure data constructed based on the node characteristics and inter-node characteristics of the AP into the graph structure model and classification model to determine the AP type, i.e., whether it is an edge AP.

[0135] Based on the same technical concept, an embodiment of the present application provides a device for handling wireless network coverage vulnerabilities, as shown in FIG8 , including:

[0136] The acquisition unit 801 is configured to acquire weak fingerprint data of each terminal within the coverage area collected by each wireless access point AP; the weak fingerprint data of any terminal is the signal data of each AP collected by the terminal under weak signal conditions;

[0137] A determining unit 802 is configured to determine a suspected coverage vulnerability based on the weak fingerprint data of each terminal;

[0138] The determining unit 802 is further configured to determine, for each suspected coverage vulnerability, whether the suspected coverage vulnerability meets a filtering condition; if the suspected coverage vulnerability meets the filtering condition, determine that the suspected coverage vulnerability does not belong to a wireless network coverage vulnerability implementation mode.

[0139] In an optional embodiment, the acquisition unit 801 is further configured to acquire weak fingerprint data of each terminal within the coverage range collected by each wireless access point AP; the weak fingerprint data of any terminal is signal data of each AP collected by the terminal under weak signal conditions; a suspected coverage vulnerability is determined based on the weak fingerprint data of each terminal and a problematic AP; the problematic AP is an AP whose signal strength is lower than a threshold among the APs associated with each terminal obtained from the cloud;

[0140] The determining unit 802 is further configured to determine, for each suspected coverage hole, whether the suspected coverage hole meets a filtering condition, and if the suspected coverage hole meets the filtering condition, determine that the suspected coverage hole does not belong to a wireless network coverage hole.

[0141] In an optional implementation, the filtering condition includes that collection times corresponding to the first coverage hole in at least two collection cycles belong to the same time range, and / or that the AP corresponding to the first coverage hole is an edge AP.

[0142] In an optional implementation, the determining unit 802 is configured to determine suspected coverage holes corresponding to different moments based on weak fingerprint data of each terminal within the coverage range of each AP obtained at different moments;

[0143] The determining unit 802 is configured to determine that any suspected coverage hole meets a filtering condition if the difference between the acquisition moments corresponding to at least two acquisition cycles is realized in a manner that meets a preset time difference threshold.

[0144] In an optional implementation manner, the determining unit 802 is further configured to: if it is determined that the AP corresponding to the suspected coverage hole is an edge AP, then the suspected coverage hole meets the filtering condition.

[0145] In an optional implementation, the determining unit 802 is further configured to determine whether an AP corresponding to the suspected coverage hole is an edge AP if the collection times corresponding to the suspected coverage hole in at least two collection cycles do not belong to the same time range; if the AP corresponding to the suspected coverage hole is an edge AP, determine that the suspected coverage hole meets the filtering condition;

[0146] or,

[0147] If the AP corresponding to the suspected coverage hole is not an edge AP, determine whether the collection moments corresponding to the suspected coverage hole in at least two collection cycles belong to the same time range; if the collection moments corresponding to the suspected coverage hole in at least two collection cycles belong to the same time range, determine that the suspected coverage hole meets the filtering conditions.

[0148] In an optional implementation, the determining unit 802 is configured to determine node characteristics of the AP corresponding to the suspected coverage hole and other APs; the node characteristics of any AP are used to represent the wireless network status of the AP;

[0149] The determining unit 802 is further configured to determine an inter-node feature between the AP corresponding to the suspected coverage hole and the other APs; wherein the inter-node feature is used to represent the correlation between AP nodes based on terminal roaming data;

[0150] The determining unit 802 is further configured to determine that the AP corresponding to the suspected coverage hole is an edge AP according to the node characteristics of the AP corresponding to the suspected coverage hole and the other APs and the inter-node characteristics.

[0151] In an optional implementation, the determining unit 802 is configured to construct AP graph structure data according to the node characteristics of the AP corresponding to the suspected coverage hole and the inter-node characteristics;

[0152] The determining unit 802 is configured to determine, based on the AP graph structure data, that the suspected coverage hole corresponds to an edge AP.

[0153] In an optional implementation, the determining unit 802 is configured to determine an attention coefficient between the AP corresponding to the suspected coverage hole and the other APs based on the node characteristics of the AP corresponding to the suspected coverage hole and the node characteristics of the other APs, and the node characteristics of the AP corresponding to the suspected coverage hole and the node characteristics of the other APs; and determine a fused node characteristic of the AP corresponding to the suspected coverage hole based on the attention coefficients of each adjacent AP;

[0154] The determining unit 802 is configured to determine, based on the fusion node feature, that the AP corresponding to the suspected coverage hole is an edge AP.

[0155] In an optional implementation, the determining unit 802 is further configured to input the AP graph structure data into a graph structure model to obtain a node feature vector of each AP;

[0156] The determining unit 802 is further configured to input the node feature vector of each AP into a classification model to obtain edge APs among the APs.

[0157] In an optional implementation, the node characteristics include AP mutual scanning data, roaming data and online and offline data of each terminal under the AP coverage;

[0158] The inter-node feature is the number of roaming times of the terminal between the nodes and / or the roaming direction of the terminal.

[0159] In an optional implementation, the determining unit 802 is configured to perform clustering processing on the weak fingerprint data of each terminal based on a density parameter to obtain first clusters and first outliers; each first outlier is weak fingerprint data that does not belong to any first cluster;

[0160] The determining unit 802 is configured to perform clustering processing on the first clusters and the first outliers based on the adjusted density parameter to obtain second clusters and second outliers until a clustering termination condition is met;

[0161] The determining unit 802 is configured to determine each cluster that meets the clustering termination condition as a suspected coverage hole; wherein the density parameters include the cluster domain size and the minimum number of points in the cluster; and the clustering process is based on the cosine distance of the weak fingerprint data.

[0162] In an optional implementation, the determining unit 802 is used to determine the clustering termination condition based on a comparison of outliers in at least two clustering results. If the number of outliers is equal, the clustering termination condition is met; otherwise, clustering continues.

[0163] In an optional implementation, the determining unit 802 is further configured to determine whether the number of terminals meets a threshold value, and if so, ultimately determine that the suspected coverage loophole is a wireless network coverage loophole.

[0164] Based on the same technical concept, an embodiment of the present application further provides a computing device, as shown in FIG9 , including at least one processor 901 and a memory 902 connected to the at least one processor. The embodiment of the present application does not limit the connection medium between the processor 901 and the memory 902. FIG9 takes the example of a bus connection between the processor 901 and the memory 902. Buses can be divided into address buses, data buses, control buses, etc.

[0165] In an embodiment of the present application, the memory 902 stores instructions that can be executed by at least one processor 901. By executing the instructions stored in the memory 902, the at least one processor 901 can execute the steps included in the aforementioned processing method based on wireless network coverage vulnerabilities.

[0166] Among them, the processor 901 is the control center of the computing device. It can use various interfaces and lines to connect various parts of the computing device, and realize data processing by running or executing instructions stored in the memory 902 and calling data stored in the memory 902. Optionally, the processor 901 may include one or more processing units. The processor 901 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, etc., and the modem processor mainly processes the issuance of instructions. It is understandable that the above-mentioned modem processor may not be integrated into the processor 901. In some embodiments, the processor 901 and the memory 902 may be implemented on the same chip. In some embodiments, they may also be implemented separately on independent chips.

[0167] The processor 901 can be a general-purpose processor, such as a central processing unit (CPU), a digital signal processor, an application-specific integrated circuit (ASIC), a field programmable gate array or other programmable logic device, a discrete gate or transistor logic device, or a discrete hardware component, and can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present application. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the embodiment of the method for processing wireless network coverage vulnerabilities can be directly embodied as being executed by a hardware processor, or can be executed by a combination of hardware and software modules in the processor.

[0168] The memory 902 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules. The memory 902 may include at least one type of storage medium, such as a flash memory, a hard disk, a multimedia card, a card-type memory, a random access memory (Random Access Memory, RAM), a static random access memory (Static Random Access Memory, SRAM), a programmable read-only memory (Programmable Read Only Memory, PROM), a read-only memory (Read Only Memory, ROM), an electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, EEPROM), a magnetic memory, a disk, an optical disk, etc. The memory 902 is any other medium that can be used to carry or store a desired program code in the form of an instruction or data structure and can be accessed by a computer, but is not limited thereto. The memory 902 in the embodiment of the present application can also be a circuit or any other device that can realize a storage function, for storing program instructions and / or data.

[0169] Based on the same technical concept, an embodiment of the present application also provides a computer-readable storage medium, which stores a computer program that can be executed by a computing device. When the program runs on the computing device, the computing device executes the steps of the above-mentioned processing method based on wireless network coverage vulnerabilities.

[0170] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0171] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the present application. It should be understood that each flow and / or box in the flow chart and / or block diagram, as well as the combination of the flow chart and / or box in the flow chart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions specified in one or more flow charts and / or one or more boxes in the block diagram.

[0172] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0173] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more processes in the flowchart and / or one or more boxes in the block diagram.

[0174] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A method for processing a wireless network coverage vulnerability, comprising: Obtain weak fingerprint data of terminals within the coverage area collected by the wireless access point AP within a preset time period; The weak fingerprint data of the terminal is the signal data of the AP collected by the terminal under weak signal conditions; Determining a suspected coverage vulnerability based on the weak fingerprint data of the terminal; Based on the suspected coverage vulnerability, determining whether the suspected coverage vulnerability meets a filtering condition; as well as When the suspected coverage vulnerability meets the filtering condition, it is determined that the suspected coverage vulnerability does not belong to a wireless network coverage vulnerability.

2. The method of claim 1, wherein: Before acquiring the weak fingerprint data of the terminals within the coverage area collected by the wireless access point AP within the preset time period, the method further includes: When the uplink signal strength of the first terminal is less than a preset threshold, a first AP is triggered to collect weak fingerprint data of the first terminal, where the first AP is a device of a network accessed by the first terminal.

3. The method of claim 1, wherein: The filtering condition is used to characterize that the suspected coverage vulnerability is not a wireless network coverage vulnerability from a time dimension and / or a space dimension.

4. The method of claim 1, wherein: The filtering condition includes that the collection time corresponding to the first coverage hole in at least two collection cycles is in the same time range, and / or the AP corresponding to the first coverage hole is an edge AP, wherein the first coverage hole is any suspected coverage hole.

5. The method according to any one of claims 1 to 4, wherein: Determine whether the suspected coverage vulnerability meets the filtering conditions, including: Based on the weak fingerprint data of the terminals within the AP coverage area obtained at different times, the suspected coverage loopholes corresponding to the different times are determined; and If the difference between the collection times corresponding to any suspected coverage hole in at least two collection cycles is less than a preset time difference threshold, the suspected coverage hole meets the filtering condition.

6. The method according to any one of claims 1 to 4, wherein: Whether the suspected coverage vulnerability meets the filtering conditions, including: When it is determined that the first AP corresponding to the suspected coverage hole is an edge AP, the suspected coverage hole meets the filtering condition.

7. The method according to any one of claims 1 to 4, wherein: Whether the suspected coverage vulnerability meets the filtering conditions, including: When the collection times corresponding to the suspected coverage hole in at least two collection cycles do not belong to the same time range, determining whether the first AP corresponding to the suspected coverage hole is an edge AP; when the first AP is an edge AP, determining that the suspected coverage hole meets the filtering condition; or, When the first AP is not an edge AP, determining whether the collection times corresponding to the suspected coverage hole in at least two collection cycles belong to the same time range; if the collection times corresponding to the suspected coverage hole in at least two collection cycles belong to the same time range, The moments belong to the same time range, and it is determined that the suspected coverage vulnerability meets the filtering condition.

8. The method according to any one of claims 1 to 7, wherein: Determining whether the AP corresponding to the suspected coverage hole meets the filtering condition includes: Determine the node characteristics of the first AP and the second AP corresponding to the suspected coverage hole; the node characteristics of any AP are used to represent the wireless network status of the AP; the second AP includes at least one other AP except the first AP; Determine an inter-node feature between the first AP and the second AP; wherein the inter-node feature is used to represent the association between AP nodes based on terminal roaming data; and The first AP is determined to be an edge AP according to respective node characteristics of the first AP and the second AP and the inter-node characteristics.

9. The method of claim 8, wherein: Determining that the first AP is an edge AP according to the node characteristics of the first AP and the second AP and the inter-node characteristics, includes: Constructing AP graph structure data according to the node characteristics of the first AP and the second AP and the inter-node characteristics; and The first AP is determined to be an edge AP through the AP graph structure data.

10. The method of claim 9, wherein: Determining, by using the AP graph structure data, that the AP corresponding to the suspected coverage hole is an edge AP includes: Inputting the AP graph structure data into a graph structure model to obtain a node feature vector of the AP; and The node feature vector of the AP is input into a classification model, and the first AP is obtained as an edge AP.

11. The method of claim 8, wherein: Determining, according to the node feature of the first AP and the inter-node feature, that the AP corresponding to the suspected coverage hole is an edge AP includes: determining an attention coefficient between the first AP and the second AP according to the node-to-node characteristics of the first AP and the second AP, the node characteristics of the first AP, and the node characteristics of the second AP; determining a fusion node characteristic of the first AP according to the attention coefficients of the adjacent APs; and According to the fusion node feature, it is determined that the first AP is an edge AP.

12. The method of claim 8, wherein: The node characteristics include one or more of AP mutual scanning data, roaming data of terminals under AP coverage, and online and offline data; The inter-node characteristics include the roaming times of the terminals between the nodes and / or the roaming directions of the terminals.

13. The method according to any one of claims 1 to 12, wherein: When the coverage of the AP includes multiple terminals, determining suspected coverage loopholes according to the weak fingerprint data of the multiple terminals includes: Clustering the weak fingerprint data of multiple terminals based on the density parameter to obtain one or more first clusters and at least one first abnormal point; the first abnormal point is weak fingerprint data that does not belong to any first cluster; Based on the adjusted density parameter, clustering the multiple first clusters and the first outlier point to obtain one or more second clusters and at least one second outlier point until a clustering termination condition is met; the second outlier point is weak fingerprint data that does not belong to any second cluster; and Each cluster that meets the cluster termination condition is determined as a suspected coverage hole; the density parameters include the cluster domain size and The minimum number of points in the cluster; the clustering process is based on the cosine distance of the weak fingerprint data; the adjusted density parameter is the cluster domain size redefined according to the multiple first clusters and the multiple first abnormal points and the minimum number of points in the cluster.

14. The method of claim 13, wherein: The clustering termination condition includes determining based on the comparison of abnormal points in at least two clustering results. When the abnormal points in the two clustering results are the same, if the abnormal points have not changed, the clustering termination condition is met.

15. The method of claim 1, wherein: After determining that the suspected coverage loophole meets the filtering condition, the method further includes: determining whether the number of terminals meets a threshold, and if so, finally determining that the suspected coverage loophole is a wireless network coverage loophole.

16. The method of claim 1, wherein: After determining whether the suspected coverage vulnerability meets the filtering condition, the method further includes: When the suspected coverage vulnerability does not meet the filtering condition, it is determined that the suspected coverage vulnerability belongs to a wireless network coverage vulnerability.

17. A device for identifying wireless network coverage loopholes, comprising: An acquisition unit is configured to acquire weak fingerprint data of each terminal within the coverage area collected by each wireless access point AP within a preset time period; The weak fingerprint data of any terminal is the signal data of each AP collected by the terminal under weak signal conditions; a determining unit configured to determine a suspected coverage vulnerability according to the weak fingerprint data of each terminal; and The determination unit is further configured to determine, for each suspected coverage vulnerability, whether the suspected coverage vulnerability meets a filtering condition; if the suspected coverage vulnerability meets the filtering condition, determine that the suspected coverage vulnerability does not belong to a wireless network coverage vulnerability.

18. A computing device comprising: A memory for storing program instructions; A processor is used to call the program instructions stored in the memory and execute the method according to any one of claims 1 to 16 according to the obtained program.

19. A computer-readable storage medium storing a computer program, wherein when the computer program is executed, the method according to any one of claims 1 to 16 is executed.

Images