A NB-IoT abnormal data testing method and system based on edge computing

Abstract

This invention discloses an edge computing-based NB-IoT anomaly data testing method and system, belonging to the field of 5G communication technology. This method improves the speed of testing NB-IoT network anomaly data and enhances the security of terminal devices by processing data on edge servers. The test terminal responds to authentication requests in various situations. Simultaneously, a multi-terminal testing method can test the specific conditions of multiple terminals in a short period of time, saving time and accelerating testing.

Description

Technical Field

[0001] The present invention belongs to the field of 5G communication technology, and in particular relates to an NB-IoT abnormal data testing method and system based on edge computing. Background Art

[0002] With the recent economic and technological development and the widespread adoption of information technology, the internet has played an increasingly important role in all areas of society and has become a strategic key to national development. Today's society places higher demands on the internet, such as massive connectivity, low cost, low energy consumption, and high stability, which existing technologies cannot meet. Narrowband Internet of Things (NB-IoT) technology is considered one of the technologies that can meet these demands and address these challenges. NB-IoT allows terminal devices to transmit data using NB-IoT base stations over narrowband networks. It utilizes various technologies, such as narrow bandwidth and spectrum boosting, to enhance network coverage and increase the number of connected devices. Current research in application areas focuses on smart parking and smart meter reading. However, due to its simple structure, NB-IoT technology also has its own limitations. For example, external attackers can exploit anomalous data to attack terminal devices operating on the NB-IoT network, thereby damaging and stealing data. Therefore, ensuring the security of NB-IoT networks and devices is a challenge. Therefore, in certain situations, it is necessary to perform anomaly detection in network traffic data.

[0003] However, current technologies still face numerous challenges and limitations when faced with increasingly complex network environments. Network anomaly data testing, as a proactive security technology, can analyze network data from terminal devices, identifying normal and abnormal network behavior, and effectively screening out problematic devices within the network. Traditional anomaly detection methods utilize a centralized computing model, processing large amounts of network data to a central server. While centralized detection methods can detect network anomaly data within a short period of time, as network data volume and complexity increase, the difficulty and processing time of anomaly detection increase significantly, making centralized computing models difficult to cope with. Furthermore, current detection systems can only detect one type of NB-IoT device. If a large number of different types of NB-IoT devices exist, repeated testing is required, increasing the workload. Summary of the Invention

[0004] In order to adapt to current network conditions and detect network traffic anomalies efficiently and accurately, the present invention proposes an NB-IoT abnormal data testing solution based on edge computing.

[0005] Edge computing technology involves splitting tasks originally performed by central servers and distributing them to individual edge servers. These edge servers are deployed close to terminals or user devices and can directly provide computing resources to them. Edge computing not only improves task processing efficiency and reduces the computing pressure on central servers in the core network, but also enhances resource utilization and security. Therefore, in narrowband IoT networks, edge computing can be used to improve the test speed and security of test systems.

[0006] The first aspect of the present invention discloses a method for testing abnormal data in NB-IoT based on edge computing. The NB-IoT is a narrowband Internet of Things (NB-IoT) comprising a central decision-making layer, a cloud center layer, an edge layer, and a device layer; wherein: a PC server is deployed in the central decision-making layer, a core network is deployed in the cloud center layer, a test data server and L edge servers are deployed in the upper layer of the edge layer, M NB-IoT platforms are deployed in the lower layer of the edge layer, and N NB-IoT devices are deployed in the device layer, where L≤M≤N; communication between the central decision-making layer and the cloud center layer, between the cloud center layer and the edge layer, and between the upper and lower layers of the edge layer is carried out using high-speed Ethernet, and communication between the edge layer and the device layer, and between the test data server and the device layer is carried out using radio frequency induction.

[0007] The method comprises:

[0008] Step S1: The PC server sends a test request, which is received by the test data server via the core network. The test request includes unique identification codes Di of P NB-IoT devices to be tested, i∈{1,2,...,P};

[0009] Step S2: The test data server classifies the P NB-IoT devices to be tested based on the test data type according to the unique identification code Di, obtains C categories, generates C types of test data corresponding to the C categories, and forwards the C types of test data to the corresponding NB-IoT devices to be tested through the radio frequency induction method, where C≤P;

[0010] Step S3: After receiving the respective test data, the P NB-IoT devices to be tested parse the test data and generate interactive communication traffic by interacting with the test data server. The unique identification code Di is attached to the interactive communication traffic, and the unique identification code Di is encapsulated as a test response message. The test response message is sent to the NB-IoT platform corresponding to the NB-IoT device through the radio frequency induction method.

[0011] Step S4: The NB-IoT platform collects the test response message from the NB-IoT device associated therewith and forwards it to an edge server with an associated relationship. The edge server pre-processes the test response message locally to obtain a test result flow ti in a standard format corresponding to the test data type.

[0012] Step S5: The edge server encapsulates the test result traffic ti and the unique identification code Di as a standard format test result message {ti, Di} and sends it to the core network. The core network parses the traffic feature information Ki from the test result traffic ti and sends the feature sequence {Ki, Di} to the PC server to determine the abnormal traffic data and the corresponding NB-IoT device.

[0013] According to the method of the first aspect of the present invention, in step S1, the PC server determines the P NB-IoT devices to be tested according to the test requirements, and obtains the unique identification codes Di of the P NB-IoT devices to be tested.

[0014] According to the method of the first aspect of the present invention, different test data types correspond to different test data, different NB-IoT devices parse the same category of test data in different ways, and the generated interactive communication traffic is also different; the PC server end located at the central decision-making layer, the test data server located at the upper layer of the edge layer, and the L edge servers all locally maintain the correspondence between the unique identification code Di of each NB-IoT device and the test data type.

[0015] According to the method of the first aspect of the present invention, in step S2, the test data server determines the type of test data corresponding to the unique identification code Di by searching based on the corresponding relationship maintained locally, obtains the C categories by classification, and forwards the C types of test data corresponding to the C categories to the corresponding NB-IoT device to be tested, where the test data is derived from the KDDCUP99 dataset or the standard NB-IoT dataset.

[0016] According to the method of the first aspect of the present invention, each edge server corresponds to several NB-IoT platforms, and each NB-IoT platform corresponds to several NB-IoT devices; in step S4, after receiving the test response message, the edge server determines the test data type corresponding to the unique identification code Di by searching based on the corresponding relationship maintained locally, and obtains the standard format corresponding to the determined test data type, and pre-processes the interactive communication traffic in the test response message according to the computing power of the edge server itself, and edits the interactive communication traffic into the test result traffic ti in the standard format, and the pre-processing includes at least: dimensionality reduction processing, noise reduction processing, data clipping, data cleaning and normalization processing.

[0017] According to the method of the first aspect of the present invention, in the step S5: the core network parses the test result traffic ti, calculates the information entropy, relative entropy and information gain of the test result traffic ti to determine its traffic feature information Ki; the PC server extracts the traffic feature information Ki and the unique identification code Di from the feature sequence {Ki, Di}, determines the test data type corresponding to the unique identification code Di by local search, and obtains the feature information threshold interval [Kmin, Kmax] corresponding to the determined test data type; if the traffic feature information Ki∈[Kmin, Kmax], the interactive communication traffic generated by the NB-IoT device is normal traffic; otherwise, it is abnormal traffic, the unique identification code Di of the NB-IoT device corresponding to the abnormal traffic is recorded, the NB-IoT device corresponding to the abnormal traffic is listed as an abnormal NB-IoT device, and an alarm is issued to the edge server and NB-IoT platform associated with the abnormal NB-IoT device.

[0018] According to the method of the first aspect of the present invention, the radio frequency induction method is to use a radio frequency induction panel to first encode the data to be transmitted on the radio frequency and then transmit it to the receiver.

[0019] The second aspect of the present invention discloses an NB-IoT abnormal data testing system based on edge computing. The system includes a central decision-making layer, a cloud center layer, an edge layer, and a device layer; wherein: a PC server is deployed in the central decision-making layer, a core network is deployed in the cloud center layer, a test data server and L edge servers are deployed in the upper layer of the edge layer, M NB-IoT platforms are deployed in the lower layer of the edge layer, and N NB-IoT devices are deployed in the device layer, where L≤M≤N; the central decision-making layer and the cloud center layer, the cloud center layer and the edge layer, and the upper and lower layers of the edge layer communicate via high-speed Ethernet, and the edge layer and the device layer, as well as the test data server and the device layer, communicate via radio frequency induction.

[0020] The system is in working condition:

[0021] The PC server sends a test request, which is received by the test data server via the core network. The test request includes unique identification codes Di of P NB-IoT devices to be tested, i∈{1,2,...,P};

[0022] The test data server classifies the P NB-IoT devices to be tested based on the test data type according to the unique identification code Di, obtains C categories and generates C types of test data corresponding to the C categories, and forwards the C types of test data to the corresponding NB-IoT devices to be tested through the radio frequency induction method, where C≤P;

[0023] After receiving the respective test data, the P NB-IoT devices to be tested parse the test data and generate interactive communication traffic by interacting with the test data server, and attach the unique identification code Di to the interactive communication traffic, encapsulate it as a test response message, and send the test response message to the NB-IoT platform corresponding to the NB-IoT device through the radio frequency induction method;

[0024] The NB-IoT platform collects the test response message from the NB-IoT device associated therewith and forwards it to an edge server with an associated relationship. The edge server pre-processes the test response message locally to obtain a test result flow ti in a standard format corresponding to the test data type;

[0025] The edge server encapsulates the test result traffic ti and the unique identification code Di as a standard format test result message {ti, Di} and sends it to the core network. The core network parses the traffic feature information Ki from the test result traffic ti and sends the feature sequence {Ki, Di} to the PC server to determine the abnormal traffic data and the corresponding NB-IoT device.

[0026] According to the system of the second aspect of the present invention, the PC server determines the P NB-IoT devices to be tested according to the test requirements, and obtains the unique identification codes Di of the P NB-IoT devices to be tested.

[0027] According to the system of the second aspect of the present invention, different test data types correspond to different test data, different NB-IoT devices parse the same category of test data in different ways, and the generated interactive communication traffic is also different; the PC server end located at the central decision-making layer, the test data server located at the upper layer of the edge layer, and the L edge servers all locally maintain the correspondence between the unique identification code Di of each NB-IoT device and the test data type.

[0028] According to the system of the second aspect of the present invention, the test data server determines the type of test data corresponding to the unique identification code Di by searching based on the corresponding relationship maintained locally, obtains the C categories by classification, and forwards the C types of test data corresponding to the C categories to the corresponding NB-IoT device to be tested, where the test data is derived from the KDDCUP99 dataset or the standard NB-IoT dataset.

[0029] According to the system of the second aspect of the present invention, each edge server corresponds to several NB-IoT platforms, and each NB-IoT platform corresponds to several NB-IoT devices; in step S4, after receiving the test response message, the edge server determines the test data type corresponding to the unique identification code Di by searching based on the correspondence maintained locally, and obtains the standard format corresponding to the determined test data type, and pre-processes the interactive communication traffic in the test response message according to the computing power of the edge server itself, and edits the interactive communication traffic into the test result traffic ti in the standard format, and the pre-processing includes at least: dimensionality reduction processing, noise reduction processing, data clipping, data cleaning and normalization processing.

[0030] According to the system of the second aspect of the present invention, the core network parses the test result traffic ti, calculates the information entropy, relative entropy and information gain of the test result traffic ti to determine its traffic feature information Ki; the PC server side extracts the traffic feature information Ki and the unique identification code Di from the feature sequence {Ki, Di}, determines the test data type corresponding to the unique identification code Di by local search, and obtains the feature information threshold interval [Kmin, Kmax] corresponding to the determined test data type; if the traffic feature information Ki∈[Kmin, Kmax], the interactive communication traffic generated by the NB-IoT device is normal traffic; otherwise, it is abnormal traffic, and the unique identification code Di of the NB-IoT device corresponding to the abnormal traffic is recorded, and the NB-IoT device corresponding to the abnormal traffic is listed as an abnormal NB-IoT device, and an alarm is issued to the edge server and NB-IoT platform associated with the abnormal NB-IoT device.

[0031] According to the system of the second aspect of the present invention, the radio frequency induction method is to use a radio frequency induction panel to first encode the data to be transmitted on the radio frequency and then transmit it to the receiver.

[0032] A third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the NB-IoT abnormal data testing method based on edge computing described in the first aspect of the present invention.

[0033] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the NB-IoT abnormal data testing method based on edge computing described in the first aspect of the present invention.

[0034] In summary, in the technical solution of the present invention, narrowband Internet of Things technology is combined with edge computing technology, and data processing is performed by the edge server to improve the testing speed of network abnormal data of the narrowband Internet of Things and improve the security of terminal equipment use. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0036] Figure 1This is a diagram showing the network composition of NB-IoT according to an embodiment of the present invention;

[0037] Figure 2 1. A diagram showing the composition of an NB-IoT abnormal data testing system according to an embodiment of the present invention;

[0038] Figure 3 FIG. 4 is a structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0039] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0040] Taking into account the limited processing capabilities of NB-IoT terminal devices, the present invention combines narrowband Internet of Things technology with edge computing technology, and performs data processing through edge servers to improve the testing speed of narrowband Internet of Things network abnormal data and improve the security of terminal device use. At the same time, the present invention can test the required NB-IoT terminal devices according to the unique identification code of the device at the same time. In the case of multiple NB-IoT terminal devices in the system, different types of NB-IoT terminal devices respond differently to the test data, and the basis for judging whether the test data is abnormal data is also different. Using the edge server, different judgment bases can be established, so that the test system can judge whether the test data is abnormal data for multiple NB-IoT terminal devices at the same time. At the same time, under certain circumstances, it is also possible to judge whether the NB-IoT terminal device has been tampered with and other information based on the response to the test data.

[0041] In view of the shortcomings of the existing technology, the present invention proposes an NB-IoT abnormal data testing solution based on edge computing to solve the defects and deficiencies of the existing technology, which is used to study the detection of abnormal data from the outside world in the NB-IoT network. Generally speaking, the amount of network data detected is relatively large and the complexity of the data is relatively high. When there are a large number of terminal devices in the network that need communication testing, the traditional centralized data processing method can no longer meet the needs of rapid detection. Therefore, by utilizing the characteristics of edge computing technology close to the data source, it can effectively reduce the data transmission delay, reduce the computing pressure of the central network, and improve the detection efficiency. By using edge computing technology to pre-process data, the same type of judgment can also be performed on different terminal devices in the system. While meeting the transmission quality of the narrowband Internet of Things, its security is improved, thereby solving the problems of narrowband Internet of Things being attacked and sensitive data being stolen to a certain extent.

[0042] The first aspect of the present invention discloses a NB-IoT abnormal data testing method based on edge computing. Figure 1 As shown, the NB-IoT is a narrowband Internet of Things, which includes a central decision-making layer, a cloud center layer, an edge layer and a device layer; wherein: a PC server is deployed in the central decision-making layer, a core network is deployed in the cloud center layer, a test data server and L edge servers are deployed in the upper layer of the edge layer, M NB-IoT platforms are deployed in the lower layer of the edge layer, and N NB-IoT devices are deployed in the device layer, L≤M≤N; the central decision-making layer and the cloud center layer, between the cloud center layer and the edge layer, and between the upper and lower layers of the edge layer all communicate using high-speed Ethernet, and the edge layer and the device layer, as well as the test data server and the device layer, all communicate using radio frequency induction.

[0043] The method comprises:

[0044] Step S1: The PC server sends a test request, which is received by the test data server via the core network. The test request includes unique identification codes Di of P NB-IoT devices to be tested, i∈{1,2,...,P};

[0045] Step S2: The test data server classifies the P NB-IoT devices to be tested based on the test data type according to the unique identification code Di, obtains C categories, generates C types of test data corresponding to the C categories, and forwards the C types of test data to the corresponding NB-IoT devices to be tested through the radio frequency induction method, where C≤P;

[0046] Step S3: After receiving the respective test data, the P NB-IoT devices to be tested parse the test data and generate interactive communication traffic by interacting with the test data server. The unique identification code Di is attached to the interactive communication traffic, and the unique identification code Di is encapsulated as a test response message. The test response message is sent to the NB-IoT platform corresponding to the NB-IoT device through the radio frequency induction method.

[0047] Step S4: The NB-IoT platform collects the test response message from the NB-IoT device associated therewith and forwards it to an edge server with an associated relationship. The edge server pre-processes the test response message locally to obtain a test result flow ti in a standard format corresponding to the test data type.

[0048] Step S5: The edge server encapsulates the test result traffic ti and the unique identification code Di as a standard format test result message {ti, Di} and sends it to the core network. The core network parses the traffic feature information Ki from the test result traffic ti and sends the feature sequence {Ki, Di} to the PC server to determine the abnormal traffic data and the corresponding NB-IoT device.

[0049] In some embodiments, in step S1, the PC server determines the P NB-IoT devices to be tested according to the test requirements, and obtains the unique identification codes Di of the P NB-IoT devices to be tested.

[0050] In some embodiments, different test data types correspond to different test data, different NB-IoT devices parse the same category of test data in different ways, and the generated interactive communication traffic is also different; the PC server end located at the central decision-making layer, the test data server located at the upper layer of the edge layer, and the L edge servers all locally maintain the correspondence between the unique identification code Di of each NB-IoT device and the test data type.

[0051] In some embodiments, in step S2, the test data server determines the test data type corresponding to the unique identification code Di by searching based on the corresponding relationship maintained locally, obtains the C categories by classification, and forwards the C types of test data corresponding to the C categories to the corresponding NB-IoT device to be tested, and the test data originates from the KDDCUP99 dataset or the standard NB-IoT dataset.

[0052] In some embodiments, each edge server corresponds to several NB-IoT platforms, and each NB-IoT platform corresponds to several NB-IoT devices; in step S4, after receiving the test response message, the edge server determines the test data type corresponding to the unique identification code Di by searching based on the correspondence maintained locally, and obtains the standard format corresponding to the determined test data type, and pre-processes the interactive communication traffic in the test response message according to the computing power of the edge server itself, and edits the interactive communication traffic into the test result traffic ti in the standard format, and the pre-processing includes at least: dimensionality reduction processing, noise reduction processing, data clipping, data cleaning and normalization processing.

[0053] In some embodiments, in step S5: the core network parses the test result traffic ti, calculates the information entropy, relative entropy and information gain of the test result traffic ti to determine its traffic feature information Ki; the PC server extracts the traffic feature information Ki and the unique identification code Di from the feature sequence {Ki, Di}, determines the test data type corresponding to the unique identification code Di by local search, and obtains the feature information threshold interval [Kmin, Kmax] corresponding to the determined test data type; if the traffic feature information Ki∈[Kmin, Kmax], the interactive communication traffic generated by the NB-IoT device is normal traffic; otherwise, it is abnormal traffic, and the unique identification code Di of the NB-IoT device corresponding to the abnormal traffic is recorded, and the NB-IoT device corresponding to the abnormal traffic is listed as an abnormal NB-IoT device, and an alarm is issued to the edge server and NB-IoT platform associated with the abnormal NB-IoT device.

[0054] In some embodiments, the radio frequency sensing method is to use a radio frequency sensing panel to first encode the data to be transmitted on the radio frequency and then transmit it to the receiver.

[0055] Specifically, the NB-IoT (Narrowband Internet of Things) is divided into four layers: the central decision-making layer, the cloud center layer, the edge layer, and the device layer.

[0056] (1) The central decision-making layer is the main body of the test system and issues test instructions when there is a test requirement. This layer stores the characteristic values ​​of normal data and data after communication between NB-IoT devices, and ultimately plays the role of judging abnormal data through these characteristic values. In general, the central decision-making layer can be a PC, and in some cases, it can also be a smartphone.

[0057] (2) The cloud center layer has powerful computing capabilities, but is far away from the data source terminal (i.e., NB-IoT device). This layer is usually used to forward, analyze and store data.

[0058] (3) The edge layer contains many edge servers and NB-IoT platforms, which include necessary facilities for network communication, such as gateways and base stations. These edge servers are usually located near the data source terminals and have certain computing and data storage capabilities. At this layer, simple data analysis and processing can be performed; the NB-IoT platform can communicate with NB-IoT devices, collect and forward necessary data; the test data server is a type of edge server, used to analyze data from the core network and generate specific test data traffic;

[0059] (4) The device layer contains a large number of NB-IoT devices with very limited computing power, which act as test terminals and communicate with the test data server. The different layers in the system framework are connected in sequence. The central decision layer, cloud center layer, and edge layer communicate via high-speed Ethernet ports, while the edge layer and device layer communicate via radio frequency induction between the NB-IoT platform and NB-IoT devices.

[0060] Specifically, the specific operation process of the abnormal data testing method of the NB-IoT network is as follows:

[0061] The PC issues a test command and sends it to the test data server: The PC at the central decision-making layer issues a test request command based on demand. After being forwarded by the core network, the test request command is sent to the test data server. The test request command carries the unique identification code Di of the NB-IoT device to be tested.

[0062] Generate test data traffic and forward it to the NB-IoT device: After the test server at the edge layer receives the test request instruction from the PC, it generates a corresponding amount of test data traffic based on the number of unique identification codes Di, and forwards it to the NB-IoT device corresponding to the unique identification code Di through radio frequency induction, and communicates with the corresponding NB-IoT device.

[0063] Generate a test communication message and forward it to the NB-IoT platform via RF induction: After receiving the test data traffic from the test server, each NB-IoT device at the device layer calculates the corresponding interactive communication traffic and retains the NB-IoT device's unique identification code Di within the interactive communication traffic to form the final test communication message. The test communication message is then forwarded to the NB-IoT platform via RF induction. The unique identification code Di can be used to distinguish NB-IoT devices.

[0064] The NB-IoT platform collects test communication messages from devices and sends them to the corresponding edge servers: The NB-IoT platform located at the edge layer collects test communication messages from different NB-IoT devices and sends them to the corresponding edge servers.

[0065] The edge server preprocesses the test communication message and forwards it to the core network: Small edge servers located at the edge layer perform test communication message preprocessing. In the field of network anomaly data testing, data preprocessing has a significant impact on anomaly detection performance. When the NB-IoT devices in the tested system are of different types, the results obtained from interacting with the test traffic will also vary. Therefore, data preprocessing in this case can significantly improve detector performance and data processing. Each edge server performs data preprocessing on the test communication message based on its own computing power, such as data dimensionality reduction. The corresponding test result traffic ti is then obtained and forwarded to the core network. When forwarding the test result traffic ti, the corresponding unique identification code Di in the test communication message is also forwarded. For example, the forwarding sequence {ti, Di}.

[0066] The core network performs feature information analysis and forwards the results to the PC. The core network analyzes the received sequence {ti, Di} based on the test traffic volume ti. Feature information extraction is not unique. In some test cases, data entropy, relative entropy, information gain, and other methods can be used to extract feature information. The corresponding feature information can be represented by Ki. The central network also forwards the analysis results to the PC. The results include the Ki value and the device's unique identification code Di.

[0067] The PC makes its final judgment: After receiving the results from the central network, it determines whether the results are within the threshold. The PC stores the characteristic information of the corresponding communication messages obtained from several normal data flows. The characteristic information obtained by analyzing several normal data flows after passing through the same NB-IoT device forms a threshold interval, such as [Kmin, Kmax]. If the characteristic value is within this interval, it is normal data flow. The PC then determines whether Ki is within the interval {Kmin, Kmax}. If not, the corresponding test data flow is considered abnormal data. The system determines that the corresponding NB-IoT device has received abnormal data by judging the characteristic information and the unique identification code, and the system issues an alarm.

[0068] The second aspect of the present invention discloses an NB-IoT abnormal data testing system based on edge computing. The system includes a central decision-making layer, a cloud center layer, an edge layer, and a device layer; wherein: a PC server is deployed in the central decision-making layer, a core network is deployed in the cloud center layer, a test data server and L edge servers are deployed in the upper layer of the edge layer, M NB-IoT platforms are deployed in the lower layer of the edge layer, and N NB-IoT devices are deployed in the device layer, where L≤M≤N; the central decision-making layer and the cloud center layer, the cloud center layer and the edge layer, and the upper and lower layers of the edge layer communicate via high-speed Ethernet, and the edge layer and the device layer, as well as the test data server and the device layer, communicate via radio frequency induction.

[0069] The system is in working condition:

[0070] The PC server sends a test request, which is received by the test data server via the core network. The test request includes unique identification codes Di of P NB-IoT devices to be tested, i∈{1,2,...,P};

[0071] The test data server classifies the P NB-IoT devices to be tested based on the test data type according to the unique identification code Di, obtains C categories and generates C types of test data corresponding to the C categories, and forwards the C types of test data to the corresponding NB-IoT devices to be tested through the radio frequency induction method, where C≤P;

[0072] After receiving the respective test data, the P NB-IoT devices to be tested parse the test data and generate interactive communication traffic by interacting with the test data server, and attach the unique identification code Di to the interactive communication traffic, encapsulate it as a test response message, and send the test response message to the NB-IoT platform corresponding to the NB-IoT device through the radio frequency induction method;

[0073] The NB-IoT platform collects the test response message from the NB-IoT device associated therewith and forwards it to an edge server with an associated relationship. The edge server pre-processes the test response message locally to obtain a test result flow ti in a standard format corresponding to the test data type;

[0074] The edge server encapsulates the test result traffic ti and the unique identification code Di as a standard format test result message {ti, Di} and sends it to the core network. The core network parses the traffic feature information Ki from the test result traffic ti and sends the feature sequence {Ki, Di} to the PC server to determine the abnormal traffic data and the corresponding NB-IoT device.

[0075] In some embodiments, the PC server determines the P NB-IoT devices to be tested according to the test requirements, and obtains the unique identification codes Di of the P NB-IoT devices to be tested.

[0076] In some embodiments, different test data types correspond to different test data, different NB-IoT devices parse the same category of test data in different ways, and the generated interactive communication traffic is also different; the PC server end located at the central decision-making layer, the test data server located at the upper layer of the edge layer, and the L edge servers all locally maintain the correspondence between the unique identification code Di of each NB-IoT device and the test data type.

[0077] In some embodiments, the test data server determines the test data type corresponding to the unique identification code Di by searching based on the corresponding relationship maintained locally, obtains the C categories by classification, and forwards the C types of test data corresponding to the C categories to the corresponding NB-IoT device to be tested, and the test data originates from the KDDCUP99 data set or the standard NB-IoT data set.

[0078] In some embodiments, each edge server corresponds to several NB-IoT platforms, and each NB-IoT platform corresponds to several NB-IoT devices; in step S4, after receiving the test response message, the edge server determines the test data type corresponding to the unique identification code Di by searching based on the correspondence maintained locally, and obtains the standard format corresponding to the determined test data type, and pre-processes the interactive communication traffic in the test response message according to the computing power of the edge server itself, and edits the interactive communication traffic into the test result traffic ti in the standard format, and the pre-processing includes at least: dimensionality reduction processing, noise reduction processing, data clipping, data cleaning and normalization processing.

[0079] In some embodiments, the core network parses the test result traffic ti, calculates the information entropy, relative entropy and information gain of the test result traffic ti to determine its traffic feature information Ki; the PC server extracts the traffic feature information Ki and the unique identification code Di from the feature sequence {Ki, Di}, determines the test data type corresponding to the unique identification code Di by local search, and obtains the feature information threshold interval [Kmin, Kmax] corresponding to the determined test data type; if the traffic feature information Ki∈[Kmin, Kmax], the interactive communication traffic generated by the NB-IoT device is normal traffic; otherwise, it is abnormal traffic, and the unique identification code Di of the NB-IoT device corresponding to the abnormal traffic is recorded, and the NB-IoT device corresponding to the abnormal traffic is listed as an abnormal NB-IoT device, and an alarm is issued to the edge server and NB-IoT platform associated with the abnormal NB-IoT device.

[0080] In some embodiments, the radio frequency sensing method is to use a radio frequency sensing panel to first encode the data to be transmitted on the radio frequency and then transmit it to the receiver.

[0081] Specifically, in some other embodiments of the abnormal data testing system of the NB-IoT network, the following modules are applied to the edge layer and the central decision layer (such as Figure 2 shown):

[0082] (1) Test request generation module 301. This module is located in the central decision-making layer and can issue test request instructions to the test system based on test requirements. The test request instruction includes the unique identification number corresponding to the NB-IoT device. For a network system with multiple NB-IoT devices, on-demand testing can be performed without any impact on NB-IoT devices that do not need testing.

[0083] (2) Test traffic generation module 302. This module is located in the test data server in the edge layer. The test data server is essentially a small network server that contains necessary communication facilities such as base stations and gateways, and also contains radio frequency sensing circuits that can communicate with NB-IoT devices. After receiving the test request instruction from the PC, the module analyzes the instruction and generates the corresponding test data traffic based on the test data set. The choice of test set is not unique. In some test instances, the KDDCUP99 dataset and / or the N-BaIoT dataset can be selected as needed.

[0084] (3) Forwarding module 303. The forwarding module is used to forward test request instructions, test data traffic, interactive communication messages and characteristic value sequences. It exists in all network layers and can include high-speed Ethernet ports and radio frequency sensing circuits.

[0085] (4) Message processing module 304. This module is located in the edge layer and the cloud center layer. It analyzes and processes the test communication message. When a test communication message is received, the edge server first pre-processes the test communication message and then passes it to the core network for feature information analysis. Data pre-processing reduces the computing load of the core network, improves the overall detection performance of the system, and greatly increases the utilization efficiency of the resources within the system. The extraction of feature information is not unique. In some test instances, the entropy value, relative entropy value, information gain, etc. of the calculated data can be selected to extract feature information.

[0086] (5) Judgment module 305. This module is located in the central decision-making layer. It is usually a judgment program in a PC. It can be written in C#. This module also stores the characteristic information of normal data. This information can be updated as the NB-IoT device and test data set change.

[0087] A third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the NB-IoT abnormal data testing method based on edge computing described in the first aspect of the present invention.

[0088] Figure 3 FIG. 1 is a structural diagram of an electronic device according to an embodiment of the present invention; FIG. Figure 3 As shown, the electronic device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the electronic device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, near field communication (NFC) or other technologies. The display screen of the electronic device can be a liquid crystal display or an electronic ink display screen, and the input device of the electronic device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the electronic device housing, or an external keyboard, touchpad or mouse.

[0089] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a structural diagram of the part related to the technical solution of the present disclosure, and does not constitute a limitation on the electronic device to which the solution of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0090] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the NB-IoT abnormal data testing method based on edge computing described in the first aspect of the present invention.

[0091] In summary, in the technical solution of the present invention, narrowband Internet of Things technology is combined with edge computing technology, and data processing is performed by the edge server to improve the testing speed of network abnormal data of the narrowband Internet of Things and improve the security of terminal equipment use.

[0092] Please note that the various technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the various technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification. The above-mentioned embodiments only express several implementation methods of the present application. The description is relatively specific and detailed, but it cannot be understood as a limitation on the scope of the invention patent. It should be pointed out that for ordinary technicians in this field, without departing from the concept of this application, several variations and improvements can be made, which all fall within the scope of protection of this application. Therefore, the scope of protection of the patent in this application shall be based on the attached claims.

Claims

1. A NB-IoT abnormal data testing method based on edge computing, characterized by: The NB-IoT is a narrowband Internet of Things, which includes a central decision-making layer, a cloud center layer, an edge layer, and a device layer; wherein: Deploy a PC server in the central decision layer, deploy a core network in the cloud center layer, deploy a test data server and L edge servers in the upper layer of the edge layer, deploy M NB-IoT platforms in the lower layer of the edge layer, and deploy N NB-IoT devices in the device layer, where L≤M≤N; The central decision layer and the cloud center layer, the cloud center layer and the edge layer, and the upper and lower layers of the edge layer all communicate via high-speed Ethernet, and the edge layer and the device layer, as well as the test data server and the device layer, all communicate via radio frequency induction. The method comprises: Step S1: The PC server sends a test request, which is received by the test data server via the core network. The test request includes unique identification codes Di of P NB-IoT devices to be tested, i∈{1,2,...,P}; Step S2: The test data server classifies the P NB-IoT devices to be tested based on the test data type according to the unique identification code Di, obtains C categories, generates C types of test data corresponding to the C categories, and forwards the C types of test data to the corresponding NB-IoT devices to be tested through the radio frequency induction method, where C≤P; Step S3: After receiving the respective test data, the P NB-IoT devices to be tested parse the test data and generate interactive communication traffic by interacting with the test data server. The unique identification code Di is attached to the interactive communication traffic, and the unique identification code Di is encapsulated as a test response message. The test response message is sent to the NB-IoT platform corresponding to the NB-IoT device through the radio frequency induction method. Step S4: The NB-IoT platform collects the test response message from the NB-IoT device associated therewith and forwards it to an edge server with an associated relationship. The edge server pre-processes the test response message locally to obtain a test result flow ti in a standard format corresponding to the test data type. Step S5: The edge server encapsulates the test result traffic ti and the unique identification code Di as a standard format test result message {ti, Di} and sends it to the core network. The core network parses the traffic feature information Ki from the test result traffic ti and sends the feature sequence {Ki, Di} to the PC server to determine the abnormal traffic data and the corresponding NB-IoT device.

2. The NB-IoT abnormal data testing method based on edge computing according to claim 1 is characterized in that: In the step S1, the PC server determines the P NB-IoT devices to be tested according to the test requirements, and obtains the unique identification codes Di of the P NB-IoT devices to be tested.

3. The NB-IoT abnormal data testing method based on edge computing according to claim 2 is characterized in that: in: Different test data types correspond to different test data. Different NB-IoT devices parse the same type of test data in different ways, and thus generate different interactive communication traffic. The PC server located at the central decision layer, the test data server located at the upper layer of the edge layer, and the L edge servers all locally maintain the correspondence between the unique identification code Di of each NB-IoT device and the test data type.

4. The NB-IoT abnormal data testing method based on edge computing according to claim 3 is characterized in that: In step S2, the test data server determines the test data type corresponding to the unique identification code Di by searching based on the corresponding relationship maintained locally, obtains the C categories by classification, and forwards the C types of test data corresponding to the C categories to the corresponding NB-IoT device to be tested, where the test data is derived from the KDDCUP99 data set or the standard NB-IoT data set.

5. The NB-IoT abnormal data testing method based on edge computing according to claim 4 is characterized in that: Each edge server corresponds to several NB-IoT platforms, and each NB-IoT platform corresponds to several NB-IoT devices; in step S4, after receiving the test response message, the edge server determines the test data type corresponding to the unique identification code Di by searching based on the correspondence maintained locally, and obtains the standard format corresponding to the determined test data type, and preprocesses the interactive communication traffic in the test response message according to the computing power of the edge server itself, and edits the interactive communication traffic into the test result traffic ti in the standard format, and the preprocessing includes at least: dimensionality reduction processing, noise reduction processing, data clipping, data cleaning and normalization processing.

6. The NB-IoT abnormal data testing method based on edge computing according to claim 5 is characterized in that: In step S5: The core network analyzes the test result traffic ti, calculates the information entropy, relative entropy and information gain of the test result traffic ti to determine its traffic characteristic information Ki; The PC server extracts the traffic feature information Ki and the unique identification code Di from the feature sequence {Ki, Di}, determines the test data type corresponding to the unique identification code Di by searching locally, and obtains the feature information threshold interval [Kmin, Kmax] corresponding to the determined test data type; If the traffic characteristic information Ki∈[Kmin,Kmax], the interactive communication traffic generated by the NB-IoT device is normal traffic; Otherwise, it is the abnormal traffic. The unique identification code Di of the NB-IoT device corresponding to the abnormal traffic is recorded, the NB-IoT device corresponding to the abnormal traffic is listed as an abnormal NB-IoT device, and an alarm is issued to the edge server and NB-IoT platform associated with the abnormal NB-IoT device.

7. The NB-IoT abnormal data testing method based on edge computing according to claim 6 is characterized in that: The radio frequency induction method is to use a radio frequency induction panel to first encode the data to be transmitted on the radio frequency and then transmit it to the receiver.

8. An NB-IoT abnormal data testing system based on edge computing, wherein the NB-IoT is a narrowband Internet of Things, characterized in that: The system includes a central decision-making layer, a cloud center layer, an edge layer, and a device layer; wherein: Deploy a PC server in the central decision layer, deploy a core network in the cloud center layer, deploy a test data server and L edge servers in the upper layer of the edge layer, deploy M NB-IoT platforms in the lower layer of the edge layer, and deploy N NB-IoT devices in the device layer, where L≤M≤N; The central decision layer and the cloud center layer, the cloud center layer and the edge layer, and the upper and lower layers of the edge layer all communicate via high-speed Ethernet, and the edge layer and the device layer, as well as the test data server and the device layer, all communicate via radio frequency induction. The system is in working condition: The PC server sends a test request, which is received by the test data server via the core network. The test request includes unique identification codes Di of P NB-IoT devices to be tested, i∈{1,2,...,P}; The test data server classifies the P NB-IoT devices to be tested based on the test data type according to the unique identification code Di, obtains C categories and generates C types of test data corresponding to the C categories, and forwards the C types of test data to the corresponding NB-IoT devices to be tested through the radio frequency induction method, where C≤P; After receiving the respective test data, the P NB-IoT devices to be tested parse the test data and generate interactive communication traffic by interacting with the test data server, and attach the unique identification code Di to the interactive communication traffic, encapsulate it as a test response message, and send the test response message to the NB-IoT platform corresponding to the NB-IoT device through the radio frequency induction method; The NB-IoT platform collects the test response message from the NB-IoT device associated therewith and forwards it to an edge server with an associated relationship. The edge server pre-processes the test response message locally to obtain a test result flow ti in a standard format corresponding to the test data type; The edge server encapsulates the test result traffic ti and the unique identification code Di as a standard format test result message {ti, Di} and sends it to the core network. The core network parses the traffic feature information Ki from the test result traffic ti and sends the feature sequence {Ki, Di} to the PC server to determine the abnormal traffic data and the corresponding NB-IoT device.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the steps in the NB-IoT abnormal data testing method based on edge computing according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the NB-IoT abnormal data testing method based on edge computing according to any one of claims 1 to 7 are implemented.

Images