A wireless communication network data security classification method and system based on edge computing

Abstract

The present invention discloses a wireless communication network data security classification method and system based on edge computing, which relates to the field of data security classification. A hidden Markov model is used to construct a wireless communication network data security state transition matrix and calculate the weight value of each node, so as to evaluate the security level of each data in the wireless communication network. The present invention uses a hidden Markov model to perform a first classification on the network data in a first data set. The hidden Markov model can be fully adapted to various types of information processing, and has low computational complexity, short evaluation time and high evaluation accuracy. The data in the first data set is classified again by an association rule classification method, which has strong applicability, high classification accuracy and short classification time. After the secondary classification is completed, one of the classification results in the two classifications is selected as the final classification result set according to the specific requirements during the classification, so as to meet various requirements and improve the scope of application.

Description

Technical Field

[0001] The present invention belongs to the field of data security classification, and specifically, particularly relates to a wireless communication network data security classification method and system based on edge computing. Background Art

[0002] Wireless communication network data is information or digital content transmitted through wireless communication technology; it can include various types of data, such as text, images, audio and video, etc. Wireless communication network data is transmitted through wireless signals in wireless communication systems, covering a wide range of application areas, including mobile communications, wireless local area networks (Wi-Fi), Bluetooth, Internet of Things (IoT), etc. These applications involve communication between people, interconnection between devices and sensors, and data exchange between different systems. The protection of wireless communication network data is crucial to ensure the security and privacy of communications. Therefore, protecting the integrity, confidentiality and availability of wireless communication network data is an important issue in wireless communication systems, and various security measures and technologies need to be taken to ensure the secure transmission and storage of data.

[0003] Chinese patent CN107864168B discloses a method and system for classifying network data streams, including: collecting an information set containing multiple dimensional information of the data stream to be processed; establishing a static behavior model and a dynamic behavior model for each dimensional information in the information set, the static behavior model characterizing the value law of the dimensional information, and the dynamic behavior model characterizing the correlation between the dimensional information at previous and subsequent times; using the static behavior model and the dynamic behavior model, respectively obtaining the static model distance and the dynamic model distance between the data stream to be processed and the data streams belonging to each target application type; based on the static model distance and the dynamic model distance between the data stream to be processed and the data streams belonging to each target application type, determining the application type of the data stream to be processed.

[0004] Wireless communication networks contain massive amounts of data, which play an important role in the current information society and people's daily lives. However, the types of data on the network are numerous and disorganized. Some data may contain Trojan viruses, etc. If these data cannot be identified and classified, people will not be able to use them efficiently. In addition, the confusion of data will waste network resources, easily pose a threat to the network, and undermine network security. Summary of the invention

[0005] In response to the problems in the related technology, the present invention proposes a wireless communication network data security classification method and system based on edge computing to overcome the above-mentioned technical problems existing in the existing related technology.

[0006] In order to solve the above technical problems, the present invention is achieved through the following technical solutions:

[0007] The present invention is a wireless communication network data security classification method, comprising the following steps:

[0008] S1. Collect a wireless communication network data sample set, recorded as the first data set; divide the wireless communication network into regions to obtain a network region set; use a hidden Markov model to construct a wireless communication network data security state transition matrix, recorded as the first matrix;

[0009] S2. According to the network area set and the first data set, the network area corresponding to each data in the first data set is obtained to generate a first data network area set; a wireless communication network data node set is set, and the optimal weight value of each node in the wireless communication network data node set is calculated according to the first matrix to obtain a node optimal weight value set;

[0010] S3, setting a data evaluation level set, evaluating the security level of each data in the wireless communication network according to the node optimal weight value set, and obtaining an evaluation result set; and then classifying the wireless communication network data according to the data evaluation level set and the evaluation result set, and obtaining a first classification set;

[0011] S4, classifying the first data set again using an association rule classification method to obtain a second classification set;

[0012] S5. Set a security classification weight and a feature classification weight. When the security classification weight is greater than or equal to the feature classification weight, use the first classification set as the final classification result set; when the security classification weight is less than the feature classification weight, use the second classification set as the final classification result set;

[0013] The first data set contains text data, image data, audio data, video data, etc. First, the network data in the first data set is classified for the first time by using the hidden Markov model. The first classification is mainly based on data security, and data with different security levels are separated. The hidden Markov model can fully adapt to various types of information processing, and has low computational complexity, short evaluation time and high evaluation accuracy. The data in the first data set is classified again by the association rule classification method. This classification is based on different characteristics of the data, and data with different characteristics and different correlations are separated. The characteristics of wireless communication network data include transmission mode characteristics. , including, for example, wired propagation, wireless propagation and broadcast transmission, signal attenuation characteristics, multipath propagation characteristics, spectrum resource limitation characteristics, time-varying channel characteristics, interference and anti-interference capability characteristics and mobility support characteristics, etc. The data is classified according to the above characteristics through the association rule classification method, and the connection between the wireless communication network data is found. It has strong applicability, high classification accuracy and short classification time. After the secondary classification is completed, according to the specific needs of the classification, such as paying more attention to the security of the data or paying more attention to the characteristics of the data, one of the classification results in the two classifications will be selected as the final classification result set, which can meet various needs and improve the scope of application.

[0014] Preferably, the S1 comprises the following steps:

[0015] S11, collect wireless communication network data samples, and obtain a wireless communication network data sample set a={a1, a2, ..., a i ,...,a a′}, a i is the i-th wireless communication network data sample in the wireless communication network data sample set, a′ is the total number of samples in the wireless communication network data sample set, and a is recorded as the first data set; at the same time, the wireless communication network is divided into regions to obtain a network region set a″={a1″,a′2′,...,a i ″,...,a′ a ″ ″}, a i ″ is the ith network area in the network area set, and a″′ is the total number of areas in the network area set;

[0016] S12, constructing a wireless communication network data security state transition matrix using a hidden Markov model, recorded as a first matrix;

[0017] By collecting wireless communication network data samples, a first data set a={a1, a2, ..., a i ,...,a a′}, which provides data support for subsequent primary and secondary classifications and improves the accuracy of classification by dividing the wireless communication network into regions.

[0018] Preferably, the S12 comprises the following steps:

[0019] S121, construct the hidden Markov model b = {b1, b2, b3}, where b1 is the wireless communication network state set space, b1 = {b 11 ,b 12 ,...,b 1i ,...,b 1b′}, b 1i is the ith network state in b1, b′ is the total number of network states in b1; b2 is the wireless communication network state observation vector space, b2 = {b 21 ,b 22 ,...,b 2i ,...,b 2b″}, b 2i is the i-th observation vector in b2, b″ is the total number of observation vectors in b2; b3 is the wireless communication network data security state transition matrix, as follows:

[0020]

[0021] Where: b 3ij =b3{b 1i →b 1j} is from b 1i Status to b 1j State change, b 1j is the jth network state in b1, b 1j ∈b1;

[0022] S122, taking b3 as the wireless communication network data security state transition matrix, recorded as the first matrix;

[0023] By constructing the first matrix, the changes in various states of the wireless communication network can be quantified, providing a classification basis for the subsequent classification process.

[0024] Preferably, S2 comprises the following steps:

[0025] S21, according to the network area set a″={a1″,a′2′,...,a i ″,...,a′ a ″ ″} and the first data set a={a1,a2,...,a i ,...,a a′}, obtain the network area corresponding to each data in the first data set, and generate the first data network area set

[0026] b″′={b1″,′b2″′,...,b i ″,′...,b a ″″},b i ″′ is a in the first data set i The network area where the wireless communication network data node set c = {c1, c2, ..., c i ,...,c c′},c i is the ith network data node in the wireless communication network data node set;

[0027] S22, and according to the first matrix and the first data network area set b″′={b1″,′b2″′,...,b i ″,′...,b a ″ ″} Calculate the optimal weight value of each node in the wireless communication network data node set to obtain the optimal node weight value set;

[0028] By calculating the weight of each node in the wireless communication network, the classification efficiency is higher and the result is more accurate.

[0029] Preferably, the S22 comprises the following steps:

[0030] S221, according to the first matrix and the first data network area set b″′={b1″,′b2″′,...,b i ″,′...,b a ″″}The calculation formula for the impact of network attacks on nodes is as follows:

[0031]

[0032] Where: f ij ′ is the weight value of network attack type; χ is the total number of network threats; f ij ″ is the probability value of threat propagation;

[0033] S222, set the maximum number of iterations to h; construct a particle population matrix g′, set the particle population size corresponding to a single wireless communication network data node to g″, then g′ is as follows:

[0034]

[0035] Where: g i ' j is the jth particle in the particle population corresponding to the i-th wireless communication network data node, and gi ' j The fitness value of is taken as the jth weight value of the i-th wireless communication network data node;

[0036] S223, set the initial fitness matrix g″′ corresponding to the particle population matrix g′ as follows:

[0037]

[0038] Where: g i ' j ″ is the initial fitness value of the jth particle in the particle population corresponding to the i-th wireless communication network data node;

[0039] The fitness calculation formula is a calculation formula for the degree of impact of network attacks on nodes;

[0040] Set the initial position of each wireless communication network data node data transmission, and obtain the initial position set h′={h1′,h2′,...,h i ′,...,h c ″}, where h i ′=(h i ′1,h i ′2); h i ′1 and h i ′2 are the horizontal and vertical coordinates of the initial position of the data transmission of c1 respectively; at the same time, the threat propagation speed set of data transmission of each wireless communication network data node is set h″={h1″,h2″,...,h i ″,...,h c ″ ′}, h i ″ represents the threat propagation speed set of data transmission of the i-th wireless communication network data node, h i ″={h i ″1,h i ″2,...,h i ' j ′,...,h i ″ g″}, h i ' j ′ is the velocity value of the jth particle in the particle population corresponding to the i-th wireless communication network data node;

[0041] S224, according to the initial fitness matrix g″′, set the initial global optimal fitness set h″′={h1″,′h2″′,...,h i ″,′...,h c ″″},h i ″′=max{gi ″1′,g i ″2′,...,g i ' j ″,...,g i ″ g ″ ′}, h i ″′ is the initial global optimal fitness of the particle population corresponding to the i-th wireless communication network data node;

[0042] S225, start iteration, update the position and fitness value of each particle in each iteration, and update the global optimal fitness of the particle population corresponding to each wireless communication network data node;

[0043] S226. When the number of iterations is greater than or equal to h, stop the iteration and obtain the final global optimal fitness set l = {l1, l2, ..., l i ,...,l c′}, l i is the final global optimal fitness of the particle population corresponding to the i-th wireless communication network data node;

[0044] S227, l={l1,l2,...,l i ,...,l c′} as the optimal weight value set of the node;

[0045] The particle swarm algorithm is used in conjunction with the first matrix and the first data network area set to iteratively tune the weight value of each node in the wireless communication network, thereby improving the efficiency of weight value calculation, and the weight value of each node after optimization is more accurate, thereby making subsequent data classification more accurate.

[0046] Preferably, S3 comprises the following steps:

[0047] S31, set data evaluation level set c″′={c1″′,c′2″,...,c′ j ″ -1 ,c′ j ″,...,c′ d ″},c′ j ″ -1 and c′ j ″ are the j-1th and jth levels in the data evaluation level set, respectively, and d is the total number of levels in the data evaluation level set; at the same time, the data evaluation threshold set e″={e1″,e′2′,...,e′ j ′,...,e e ″ ″′}, e″′=d-1, e′j ′ is the jth threshold in the data evaluation threshold set, and e″′ is the total number of thresholds in the data evaluation threshold set;

[0048] S32, according to the optimal node weight value set l = {l1, l2, ..., l i ,...,l c′}Evaluate the security level of each data in the wireless communication network and obtain the evaluation result set d′={d1′,d2′,...,d i ′,...,d a ″},d i ′ is a i The results of the safety assessment;

[0049] S33, evaluating the level set c″′={c1″′, c′2″, ..., c′ according to the data j ″ -1 ,c′ j ″,...,c′ d ″} and the evaluation result set d′={d1′,d2′,...,d i ′,...,d a ″} Classify the wireless communication network data to obtain a first classification set

[0050] d″={d1″,d2″,...,d i ″,...,d′ d ″ ″}, the classification process is as follows:

[0051] when e′ j ' -1 ≤d i '<e' j ′, then d i The classification level of ′ is c′ j ″ -1 , when e e ″ ″′ ≤d i ′, then d i The classification level of ′ is c′ d ″;

[0052] By setting a data evaluation level set, a specific classification standard is provided for classifying wireless communication network data, thereby facilitating classification.

[0053] Preferably, the S32 comprises the following steps:

[0054] S321, set the data sample matrix l' corresponding to each wireless communication network data node as follows:

[0055]

[0056] Where: l i ' j is the jth wireless communication network data sample in the i-th wireless communication network data node, l1″, l2″, ..., l c ″ ′ are the total number of wireless communication network data samples in the 1st, 2nd, ..., c′th wireless communication network data nodes, respectively,

[0057] S322, calculate the evaluation result matrix l″′ according to l′, as follows:

[0058]

[0059] Where: l i ' j ″=m ij ·l i ' j , m ij is the threat level of the jth wireless communication network data sample in the ith wireless communication network data node to the ith wireless communication network data node;

[0060] S323, according to the evaluation result matrix l″′, obtain the evaluation result set d′={d1′, d2′, ..., d i ′,...,d a ″};

[0061] The security level of the data in each wireless communication network data node is calculated according to the threat level and the weight value of the wireless communication network data node, so that security classification can be performed.

[0062] Preferably, S4 comprises the following steps:

[0063] S41, using association rule classification method to classify the first data set a={a1, a2, ..., a i ,...,a a′} to mine the attribute associations between the data and obtain the feature data classification set m′={m1′,m′2,...,m i ′,...,m′ m″}, where m i ′ is the i-th category in the feature data classification set, and m″ is the total number of categories in the feature data classification set;

[0064] The specific steps include:

[0065] S411, set the a={a1, a2, ..., ai ,...,a a′ The data set with association rule attributes in} is

[0066] m″′={m1″′,m′2″,...,m i ″′,...,m′ n ″}, recorded as the second data set, n≤a′,m i ″′ is the i-th data in the second data set;

[0067] S412: The second data set m″′={m1″′,m′2″,...,m i ″′,...,m′ n ″} are binary converted to obtain a second data set n′={n1′,n′2,...,n i ′,...,n′ n}, recorded as the third data set; n i ′ is m i ″′ is converted into binary data; and then the third data set n′={n1′,n′2,...,n i ′,...,n′ n}, and obtain the attribute support rate set n″={n1″,n′2′,...,n i ″,...,n′ n ′},n i ″ is n i ′’s attribute support rate;

[0068] S413, set the minimum support rate n'', traverse the attribute support rate set and compare it with the minimum support rate n'', when the data in the attribute support rate set is greater than or equal to n'', the data is regarded as frequent item data, after the traversal is completed, the first frequent item data set is obtained o 1i is the i-th frequent item data in the first frequent item data set;

[0069] S414, cyclically scan the first frequent item data set, and obtain a new frequent item data set each time the scan is performed. It is assumed that after the o″+1th scan, no new frequent item data set is generated, and the scanning is stopped to obtain a frequent item data set matrix o″′, as follows:

[0070]

[0071] Among them: jiis the i-th frequent item data in the new frequent item data set generated after the j-th scan, o1′, o′2, ..., o′ o″ are the total number of frequent item data in the new frequent item data set obtained after the first scan, the second scan, ... the o″th scan, respectively;

[0072] S42, obtaining a second classification set e={e1, e2, ..., e j ,...,e o″},in e j is the jth category in the second category set;

[0073] The association rule classification method is used to classify the data according to the above characteristics and find out the connection between the wireless communication network data. It has strong applicability, high classification accuracy and short classification time.

[0074] Preferably, the S5 comprises the following steps:

[0075] S51, setting a security classification weight α and a feature classification weight β according to a wireless communication network data classification standard;

[0076] S52, when α≥β, the first classification set d″={d1″, d2″, ..., d i ″,...,d′ d ″ ″} as the final classification result set; when α<β, the second classification set e={e1,e2,...,e i ,...,e e′} as the final classification result set;

[0077] According to the specific needs during classification, such as paying more attention to data security or paying more attention to data characteristics, one of the classification results of the two classifications will be selected as the final classification result set, thereby meeting multiple needs and improving the scope of application.

[0078] The invention also discloses a system for realizing the wireless communication network data security classification method.

[0079] A wireless communication network data security classification system based on edge computing, comprising a network data sample acquisition module, a weight value calculation module, a first network data evaluation classification module, a second network data evaluation classification module and a final classification module;

[0080] The network data sample collection module is used to collect wireless communication network data sample sets;

[0081] The weight value calculation module is used to calculate the weight value of each node in the wireless communication network data node set according to the first matrix;

[0082] The first network data evaluation classification module is used to classify the wireless communication network data according to the data evaluation level set and the evaluation result set;

[0083] The second network data evaluation and classification module is used to perform secondary classification on the first data set using an association rule classification method;

[0084] The final classification module is used to obtain a final classification result set according to the first classification set and the second classification set.

[0085] The present invention has the following beneficial effects:

[0086] In the present invention, a network data sample collection module, a weight value calculation module, a first network data evaluation classification module, a second network data evaluation classification module and a final classification module are set. First, the network data in the first data set is classified for the first time by using a hidden Markov model. The first classification is mainly based on data security, and data with different security levels are separated. The hidden Markov model can be fully adapted to various types of information processing, and has low computational complexity, short evaluation time and high evaluation accuracy. The data in the first data set is classified again by an association rule classification method. This classification is based on different characteristics of the data, and data with different characteristics and different correlations are separated. The characteristics of wireless communication network data include transmission mode characteristics, including, for example, wired transmission, wireless transmission and broadcast transmission, signal attenuation characteristics, Multipath propagation characteristics, spectrum resource limitation characteristics, time-varying channel characteristics, interference and anti-interference capability characteristics and mobility support characteristics, etc., are classified according to the above characteristics through association rule classification method, and the connection between wireless communication network data is found. It has strong applicability, high classification accuracy and short classification time. After the secondary classification is completed, according to the specific needs during classification, such as paying more attention to data security or paying more attention to data characteristics, one of the classification results in the two classifications is selected as the final classification result set, so as to meet various needs and improve the scope of application. In addition, the data transmission between the network data sample acquisition module, the weight value calculation module, the first network data evaluation classification module, the second network data evaluation classification module and the final classification module in the present invention adopts asynchronous data transmission technology to improve data transmission efficiency.

[0087] In the present invention, by constructing the first matrix, the changes in various states of the wireless communication network can be quantified, providing a classification basis for the subsequent classification process and calculating the weight of each node in the wireless communication network, thereby making the classification more efficient and the results more accurate.

[0088] The present invention uses a particle swarm algorithm in conjunction with a first matrix and a first data network region set to iteratively tune the weight values ​​of each node in a wireless communication network, thereby improving the efficiency of weight value calculation, and the weight values ​​of each node after optimization are more accurate, thereby making subsequent data classification more accurate.

[0089] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] In order to more clearly illustrate the technical solutions of the embodiments of the invention, the drawings required for describing the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the invention. For ordinary technicians in this field, they can also obtain drawings based on these drawings without paying creative work.

[0091] Figure 1 A schematic diagram of a flow chart of classifying wireless communication network data by a wireless communication network data security classification system based on edge computing of the present invention. DETAILED DESCRIPTION

[0092] The following will be combined with the drawings in the embodiments of the invention to clearly and completely describe the technical solutions in the embodiments of the invention. Obviously, the described embodiments are only part of the embodiments of the invention, not all of the embodiments. Based on the embodiments in the invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the invention.

[0093] In the description of the present invention, it is necessary to understand that the terms "opening", "upper", "lower", "top", "middle", "inside" and the like indicating orientation or positional relationship are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as limiting the invention.

[0094] The present invention is a wireless communication network data security classification method, comprising the following steps:

[0095] S1. Collect a wireless communication network data sample set, recorded as the first data set; divide the wireless communication network into regions to obtain a network region set; use a hidden Markov model to construct a wireless communication network data security state transition matrix, recorded as the first matrix;

[0096] The S1 comprises the following steps:

[0097] S11, collect wireless communication network data samples, and obtain a wireless communication network data sample set a={a1, a2, ..., ai ,...,a a′}, a i is the i-th wireless communication network data sample in the wireless communication network data sample set, a′ is the total number of samples in the wireless communication network data sample set, and a is recorded as the first data set; at the same time, the wireless communication network is divided into regions to obtain a network region set a″={a1″,a′2′,...,a i ″,...,a′ a ″ ″}, a i ″ is the ith network area in the network area set, and a″′ is the total number of areas in the network area set;

[0098] S12, constructing a wireless communication network data security state transition matrix using a hidden Markov model, recorded as a first matrix;

[0099] The S12 comprises the following steps:

[0100] S121, construct the hidden Markov model b = {b1, b2, b3}, where b1 is the wireless communication network state set space, b1 = {b 11 ,b 12 ,...,b 1i ,...,b 1b′}, b 1i is the ith network state in b1, b′ is the total number of network states in b1; b2 is the wireless communication network state observation vector space, b2 = {b 21 ,b 22 ,...,b 2i ,...,b 2b″}, b 2i is the i-th observation vector in b2, b″ is the total number of observation vectors in b2; b3 is the wireless communication network data security state transition matrix, as follows:

[0101]

[0102] Where: b 3ij =b3{b 1i →b 1j} is from b 1i Status to b 1j State change, b 1j is the jth network state in b1, b 1j ∈b1;

[0103] S122, taking b3 as the wireless communication network data security state transition matrix, recorded as the first matrix;

[0104] S2. According to the network area set and the first data set, the network area corresponding to each data in the first data set is obtained to generate a first data network area set; a wireless communication network data node set is set, and the optimal weight value of each node in the wireless communication network data node set is calculated according to the first matrix to obtain a node optimal weight value set;

[0105] The S2 comprises the following steps:

[0106] S21, according to the network area set a″={a1″,a′2′,...,a i ″,...,a′ a ″ ″} and the first data set a={a1,a2,...,a i ,...,a a′}, obtain the network area corresponding to each data in the first data set, and generate the first data network area set

[0107] b″′={b1″,′b2″′,...,b i ″,′...,b a ″ ″}, b i ″′ is a in the first data set i The network area where the wireless communication network data node set c = {c1, c2, ..., c i ,...,c c′},c i is the ith network data node in the wireless communication network data node set;

[0108] S22, and according to the first matrix and the first data network area set b″′={b1″,′b2″′,...,b i ″,′...,b a Calculate the optimal weight value of each node in the wireless communication network data node set to obtain the optimal node weight value set;

[0109] The S22 comprises the following steps:

[0110] S221, according to the first matrix and the first data network area set b″′={b1″,′b2″′,...,b i ″,′...,b a ″″}The calculation formula for the impact of network attacks on nodes is as follows:

[0111]

[0112] Where: f ij ′ is the weight value of network attack type; χ is the total number of network threats; fij ″ is the probability value of threat propagation;

[0113] S222, set the maximum number of iterations to h; construct a particle population matrix g′, set the particle population size corresponding to a single wireless communication network data node to g″, then g′ is as follows:

[0114]

[0115] Where: g i ' j is the jth particle in the particle population corresponding to the i-th wireless communication network data node, and g i ' j The fitness value of is taken as the jth weight value of the i-th wireless communication network data node;

[0116] S223, set the initial fitness matrix g″′ corresponding to the particle population matrix g′ as follows:

[0117]

[0118] Where: g i ' j ″ is the initial fitness value of the jth particle in the particle population corresponding to the i-th wireless communication network data node;

[0119] The fitness calculation formula is a calculation formula for the degree of impact of network attacks on nodes;

[0120] Set the initial position of each wireless communication network data node data transmission, and obtain the initial position set h′={h1′,h2′,...,h i ′,...,h c ″}, where h i ′=(h i ′1,h i ′2); h i ′1 and h i ′2 are the horizontal and vertical coordinates of the initial position of the data transmission of c1 respectively; at the same time, the threat propagation speed set of data transmission of each wireless communication network data node is set h″={h1″,h2″,...,h i ″,...,h c ″ ′}, h i ″ represents the threat propagation speed set of data transmission of the i-th wireless communication network data node, h i ″={h i ″1,h i ″2,...,h i ' j ′,...,h i ″g″}, h i ' j ′ is the velocity value of the jth particle in the particle population corresponding to the i-th wireless communication network data node;

[0121] S224, according to the initial fitness matrix g″′, set the initial global optimal fitness set h″′={h1″,′h2″′,...,h i ″,′...,h c ″″},h i ″′=max{g i ″1′,g i ″2′,...,g i ' j ″,...,g i ″ g ″ ′}, h i ″′ is the initial global optimal fitness of the particle population corresponding to the i-th wireless communication network data node;

[0122] S225, start iteration, update the position and fitness value of each particle in each iteration, and update the global optimal fitness of the particle population corresponding to each wireless communication network data node;

[0123] S226. When the number of iterations is greater than or equal to h, stop the iteration and obtain the final global optimal fitness set l = {l1, l2, ..., l i ,...,l c′}, l i is the final global optimal fitness of the particle population corresponding to the i-th wireless communication network data node;

[0124] S227, l={l1,l2,...,l i ,...,l c′} as the optimal weight value set of the node;

[0125] S3, setting a data evaluation level set, evaluating the security level of each data in the wireless communication network according to the node optimal weight value set, and obtaining an evaluation result set; and then classifying the wireless communication network data according to the data evaluation level set and the evaluation result set, and obtaining a first classification set;

[0126] The S3 comprises the following steps:

[0127] S31. Set data evaluation level set c″′={c1″′,c′2″,...,c′ j ″-1 ,c′ j ″,...,c′ d ″},c′ j ″ -1 and c′ j ″ are the j-1th and jth levels in the data evaluation level set, respectively, and d is the total number of levels in the data evaluation level set; at the same time, the data evaluation threshold set e″={e1″,e′2′,...,e′ j ′,...,e e ″ ″′}, e″′=d-1, e′ j ′ is the jth threshold in the data evaluation threshold set, and e″′ is the total number of thresholds in the data evaluation threshold set;

[0128] S32, according to the optimal node weight value set l = {l1, l2, ..., l i ,...,l c′}Evaluate the security level of each data in the wireless communication network and obtain the evaluation result set d′={d1′,d2′,...,d i ′,...,d a ″},d i ′ is a i The results of the safety assessment;

[0129] The S32 comprises the following steps:

[0130] S321, set the data sample matrix l' corresponding to each wireless communication network data node as follows:

[0131]

[0132] Where: l i ' j is the jth wireless communication network data sample in the i-th wireless communication network data node, l1″, l2″, ..., l c ″ ′ are the total number of wireless communication network data samples in the 1st, 2nd, ..., c′th wireless communication network data nodes, respectively,

[0133] S322, calculate the evaluation result matrix l″′ according to l′, as follows:

[0134]

[0135] Where: l i ' j ″=m ij ·l i ' j, m ij is the threat level of the jth wireless communication network data sample in the ith wireless communication network data node to the ith wireless communication network data node;

[0136] S323, according to the evaluation result matrix l″′, obtain the evaluation result set d′={d1′, d2′, ..., d i ′,...,d a ″};

[0137] S33, evaluating the level set c″′={c1″′, c′2″, ..., c′ according to the data j ″ -1 ,c′ j ″,...,c′ d ″} and the evaluation result set d′={d1′,d2′,...,d i ′,...,d a ″} Classify the wireless communication network data to obtain a first classification set

[0138] d″={d1″,d2″,...,d i ″,...,d′ d ″ ″}, the classification process is as follows:

[0139] when e′ j ' -1 ≤d i '<e' j ′, then d i The classification level of ′ is c′ j ″ -1 , when e e ″ ″′ ≤d i ′, then d i The classification level of ′ is c′ d ″;

[0140] S4, classifying the first data set again using an association rule classification method to obtain a second classification set;

[0141] The S4 comprises the following steps:

[0142] S41, using association rule classification method to classify the first data set a={a1, a2, ..., a i ,...,a a′} to mine the attribute associations between the data and obtain the feature data classification set m′={m1′,m′2,...,m i ′,...,m′ m″}, where m i′ is the i-th category in the feature data classification set, and m″ is the total number of categories in the feature data classification set;

[0143] The specific steps include:

[0144] S411, set the a={a1, a2, ..., a i ,...,a a′ The data set with association rule attributes in} is

[0145] m″′={m1″′,m′2″,...,m i ″′,...,m′ n ″}, recorded as the second data set, n≤a′,m i ″′ is the i-th data in the second data set;

[0146] S412: The second data set m″′={m1″′,m′2″,...,m i ″′,...,m′ n ″} are binary converted to obtain a second data set n′={n1′,n′2,...,n i ′,...,n′ n}, recorded as the third data set; n i ′ is m i ″′ is converted into binary data; and then the third data set n′={n1′,n′2,...,n i ′,...,n′ n}, and obtain the attribute support rate set n″={n1″,n′2′,...,n i ″,...,n′ n ′},n i ″ is n i ′’s attribute support rate;

[0147] S413, set the minimum support rate n'', traverse the attribute support rate set and compare it with the minimum support rate n'', when the data in the attribute support rate set is greater than or equal to n'', the data is regarded as frequent item data, after the traversal is completed, the first frequent item data set o1 = {o 11 ,o 12 ,...,o 1i ,...,o 1o1′},o 1i is the i-th frequent item data in the first frequent item data set;

[0148] S414, cyclically scan the first frequent item data set, and obtain a new frequent item data set each time the scan is performed. It is assumed that after the o″+1th scan, no new frequent item data set is generated, and the scanning is stopped to obtain a frequent item data set matrix o″′, as follows:

[0149]

[0150] Among them: ji is the i-th frequent item data in the new frequent item data set generated after the j-th scan, o1′, o′2, ..., o′ o″ are the total number of frequent item data in the new frequent item data set obtained after the first scan, the second scan, ... the o″th scan, respectively;

[0151] S42, obtaining a second classification set e={e1, e2, ..., e j ,...,e o″},in

[0152] e j is the jth category in the second category set;

[0153] S5. Set a security classification weight and a feature classification weight. When the security classification weight is greater than or equal to the feature classification weight, use the first classification set as the final classification result set; when the security classification weight is less than the feature classification weight, use the second classification set as the final classification result set;

[0154] The S5 comprises the following steps:

[0155] S51, setting a security classification weight α and a feature classification weight β according to a wireless communication network data classification standard;

[0156] S52, when α≥β, the first classification set d″={d1″, d2″, ..., d i ″,...,d′ d ″ ″} as the final classification result set; when α<β, the second classification set e={e1,e2,...,e i ,...,e e′} as the final classification result set.

[0157] A wireless communication network data security classification system based on edge computing, comprising a network data sample acquisition module, a weight value calculation module, a first network data evaluation classification module, a second network data evaluation classification module and a final classification module;

[0158] The network data sample collection module is used to collect wireless communication network data sample sets;

[0159] The weight value calculation module is used to calculate the weight value of each node in the wireless communication network data node set according to the first matrix;

[0160] The first network data evaluation classification module is used to classify the wireless communication network data according to the data evaluation level set and the evaluation result set;

[0161] The second network data evaluation and classification module is used to perform secondary classification on the first data set using an association rule classification method;

[0162] The final classification module is used to obtain a final classification result set according to the first classification set and the second classification set.

[0163] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0164] The preferred embodiments of the invention disclosed above are only used to help explain the invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the invention, so that those skilled in the art can understand and use the invention well.

Claims

1. A wireless communication network data security classification method, characterized in that: The following steps are involved: S1. Collect a wireless communication network data sample set, recorded as the first data set; divide the wireless communication network into regions to obtain a network region set; use a hidden Markov model to construct a wireless communication network data security state transition matrix, recorded as the first matrix; The S1 comprises the following steps: S11, collect wireless communication network data samples, and obtain a wireless communication network data sample set a={a1, a2, ..., a i ,...,a a′ }, a i is the i-th wireless communication network data sample in the wireless communication network data sample set, a′ is the total number of samples in the wireless communication network data sample set, and a is recorded as the first data set; at the same time, the wireless communication network is divided into regions to obtain a network region set a″={a1″,a′2′,...,a i ″,...,a′ a ″ ″ }, a i ″ is the ith network area in the network area set, and a″′ is the total number of areas in the network area set; S12, constructing a wireless communication network data security state transition matrix using a hidden Markov model, recorded as a first matrix; The S12 comprises the following steps: S121, construct the hidden Markov model b = {b1, b2, b3}, where b1 is the wireless communication network state set space, b1 = {b 11 ,b 12 ,...,b 1i ,...,b 1b′ }, b 1i is the ith network state in b1, b′ is the total number of network states in b1; b2 is the wireless communication network state observation vector space, b2 = {b 21 ,b 22 ,...,b 2i ,...,b 2b″ }, b 2i is the i-th observation vector in b2, b″ is the total number of observation vectors in b2; b3 is the wireless communication network data security state transition matrix, as follows: Where: b 3ij =b3{b 1i →b 1j } is from b 1i Status to b 1j State change, b 1j is the jth network state in b1, b 1j ∈b1; S122, taking b3 as the wireless communication network data security state transition matrix, recorded as the first matrix; S2. According to the network area set and the first data set, the network area corresponding to each data in the first data set is obtained to generate a first data network area set; a wireless communication network data node set is set, and the optimal weight value of each node in the wireless communication network data node set is calculated according to the first matrix to obtain an optimal node weight value set; The S2 comprises the following steps: S21, according to the network area set a″={a1″,a′2′,...,a i ″,...,a′ a ″ ″ } and the first data set a={a1,a2,...,a i ,...,a a′ }, obtain the network area corresponding to each data in the first data set, and generate the first data network area set b″′={b1″,′b2″′,...,b i ″,′...,b a ″″},b i ″′ is a in the first data set i The network area where the wireless communication network data node set c = {c1, c2, ..., c i ,...,c c′ }, c i is the ith network data node in the wireless communication network data node set; S22, and according to the first matrix and the first data network area set b″′={b1″,′b2″′,...,b i ″,′...,b a ″ ″ } Calculate the optimal weight value of each node in the wireless communication network data node set to obtain the optimal node weight value set; The S22 comprises the following steps: S221, according to the first matrix and the first data network area set b″′={b1″,′b2″′,...,b i ″,′...,b a ″″}The calculation formula for the impact of network attacks on nodes is as follows: Where: f ij ′ is the weight value of network attack type; χ is the total number of network threats; f ij ″ is the probability value of threat propagation; S222, set the maximum number of iterations to h; construct a particle population matrix g′, set the particle population size corresponding to a single wireless communication network data node to g″, then g′ is as follows: Where: g i ' j is the jth particle in the particle population corresponding to the i-th wireless communication network data node, and g i ' j The fitness value of is taken as the jth weight value of the i-th wireless communication network data node; S223, set the initial fitness matrix g″′ corresponding to the particle population matrix g′ as follows: Where: g″′ ij is the initial fitness value of the jth particle in the particle population corresponding to the i-th wireless communication network data node; The fitness calculation formula is a calculation formula for the degree of impact of network attacks on nodes; Set the initial position of data transmission of each wireless communication network data node to obtain the initial position set h′={h′1,h′2,...,h′ i ,...,h′ c′ }, where h′ i =(h′ i1 ,h′ i2 );h′ i1 and h′ i2 are the horizontal and vertical coordinates of the initial position of c1 data transmission respectively; at the same time, the threat propagation speed set h″={h″1,h″2,...,h″ i ,...,h″ c′ }, h″ i represents the threat propagation speed set of data transmission of the i-th wireless communication network data node, h″ i ={h″ i1 ,h″ i2 ,...,h″ ij ,...,h″ ig″ }, h″ ij is the velocity value of the jth particle in the particle population corresponding to the i-th wireless communication network data node; S224, according to the initial fitness matrix g″′, set the initial global optimal fitness set h″′={h″′1,h″′2,...,h″′ of the particle population corresponding to each wireless communication network data node i ,...,h″′ c′ },h″′ i =max{g″′ i1 ,g″′ i2 ,...,g″′ ij ,...,g″′ ig′ },h″′ i is the initial global optimal fitness of the particle population corresponding to the i-th wireless communication network data node; S225, start iteration, update the position and fitness value of each particle in each iteration, and update the global optimal fitness of the particle population corresponding to each wireless communication network data node; S226. When the number of iterations is greater than or equal to h, stop the iteration and obtain the final global optimal fitness set l = {l1, l2, ..., l i ,...,l c′ }, l i is the final global optimal fitness of the particle population corresponding to the i-th wireless communication network data node; S227, l={l1,l2,...,l i ,...,l c′ } as the optimal weight value set of the node; S3, setting a data evaluation level set, evaluating the security level of each data in the wireless communication network according to the node optimal weight value set, and obtaining an evaluation result set; and then classifying the wireless communication network data according to the data evaluation level set and the evaluation result set, and obtaining a first classification set; S4, classifying the first data set again using an association rule classification method to obtain a second classification set; The S4 comprises the following steps: S41, using association rule classification method to classify the first data set a={a1, a2, ..., a i ,...,a a′ } to mine the attribute associations between the data in the dataset and obtain the feature data classification set m′={m′1,m′2,...,m′ i ,...,m′ m″ }, where m′ i is the i-th category in the feature data classification set, and m″ is the total number of categories in the feature data classification set; The specific steps include: S411, set the a={a1, a2, ..., a i ,...,a a′ The data set with association rule attributes in} is m″′={m″′1,m″′2,...,m″′ i ,...,m″′ n }, recorded as the second data set, n≤a′,m″′ i is the i-th data in the second data set; S412: The second data set m″′={m″′1, m″′2, ..., m″′ i ,...,m″′ n } are binary converted to obtain a second data set n′={n′1,n′2,...,n′ i ,...,n′ n }, recorded as the third data set; n′ i m″′ i The data obtained after binary conversion; and then calculating the third data set n′={n′1, n′2, ..., n′ i ,...,n′ n }, and obtain the attribute support rate set n″={n″1,n″2,...,n″ i ,...,n″ n },n″ i is n′ i Attribute support rate; S413, set the minimum support rate n'', traverse the attribute support rate set and compare it with the minimum support rate n'', when the data in the attribute support rate set is greater than or equal to n'', the data is regarded as frequent item data, after the traversal is completed, the first frequent item data set is obtained o 1i is the i-th frequent item data in the first frequent item data set; S414, cyclically scan the first frequent item data set, and obtain a new frequent item data set each time the scan is performed. It is assumed that after the o″+1th scan, no new frequent item data set is generated, and the scanning is stopped to obtain a frequent item data set matrix o″′, as follows: Among them: ji is the i-th frequent item data in the new frequent item data set generated after the j-th scan, o′1, o′2, ..., o′ o″ are the total number of frequent item data in the new frequent item data set obtained after the first scan, the second scan, ... the o″th scan, respectively; S42, obtaining a second classification set e={e1, e2, ..., e j ,...,e o″ },in e j is the jth category in the second category set; S5. Set the security classification weight and feature classification weight. When the security classification weight is greater than or equal to the feature classification weight, use the first classification set as the final classification result set; when the security classification weight is less than the feature classification weight, use the second classification set as the final classification result set.

2. A wireless communication network data security classification method according to claim 1, characterized in that: The S3 comprises the following steps: S31. Set data evaluation level set c″′={c″′1, c″′2, ..., c″′ j-1 ,c″′ j ,...,c″′ d },c″′ j-1 and c″′ j are the j-1th and jth levels in the data evaluation level set, respectively, and d is the total number of levels in the data evaluation level set; at the same time, the data evaluation threshold set e″={e″1,e″2,...,e″ j ,...,e″ e″′ }, e″′=d-1, e″ j is the jth threshold in the data evaluation threshold set, e″′ is the total number of thresholds in the data evaluation threshold set; S32, according to the optimal node weight value set l = {l1, l2, ..., l i ,...,l c′ }Evaluate the security level of each data in the wireless communication network and obtain the evaluation result set d′={d′1,d′2,...,d′ i ,...,d′ a′ }, d′ i for a i The results of the safety assessment; S33, evaluating the level set c″′={c″′1, c″′2, ..., c″′ according to the data j-1 ,c″′ j ,...,c″′ d } and the evaluation result set d′={d′1,d′2,...,d′ i ,...,d′ a′ }Classify the wireless communication network data to obtain a first classification set d″={d″1,d″2,...,d″ i ,...,d″′ d″ }, the classification process is as follows: When e″ j-1 ≤d′ i <e″ j When d′ i The classification level is c″′ j-1 , when e″ e″′ ≤d i ′, then d′ i The classification level is c″′ d .

3. A wireless communication network data security classification method according to claim 2, characterized in that: The S32 comprises the following steps: S321, set the data sample matrix l' corresponding to each wireless communication network data node as follows: Where: l′ ij is the jth wireless communication network data sample in the i-th wireless communication network data node, l″1, l″2, ..., l″ c′ are the total number of wireless communication network data samples in the 1st, 2nd, ..., c′th wireless communication network data nodes, respectively, S322, calculate the evaluation result matrix l″′ according to l′, as follows: Where: l″′ ij =m ij l′ ij , m ij is the threat level of the jth wireless communication network data sample in the ith wireless communication network data node to the ith wireless communication network data node; S323, obtaining the evaluation result set d′={d′1, d′2, ..., d′ according to the evaluation result matrix l″′ i ,...,d′ a′ }.

4. A wireless communication network data security classification method according to claim 1, characterized in that: The S5 comprises the following steps: S51, setting a security classification weight α and a feature classification weight β according to a wireless communication network data classification standard; S52, when α≥β, the first classification set d″={d″1,d″2,...,d″ i ,...,d″ d″ } as the final classification result set; when α<β, the second classification set e={e1,e2,...,e i ,...,e e′ } as the final classification result set.

5. A wireless communication network data security classification system based on edge computing that implements the wireless communication network data security classification method according to any one of claims 1 to 4, characterized in that: It includes a network data sample collection module, a weight value calculation module, a first network data evaluation classification module, a second network data evaluation classification module and a final classification module; The network data sample collection module is used to collect wireless communication network data sample sets; The weight value calculation module is used to calculate the weight value of each node in the wireless communication network data node set according to the first matrix; The first network data evaluation classification module is used to classify the wireless communication network data according to the data evaluation level set and the evaluation result set; The second network data evaluation and classification module is used to perform secondary classification on the first data set using an association rule classification method; The final classification module is used to obtain a final classification result set according to the first classification set and the second classification set.

Images