Design of Network Security Management System Based on
K-Means Algorithm
Aiping Liu
Quanzhou University of Information Engineering, Quanzhou, China
Keywords: K-Means Algorithm, Network, Safety Management.
Abstract: The conventional network security management system focuses on the security of network data storage, and
does not establish a network management architecture according to the actual needs of the system, which
makes the network management always have security risks. Therefore, a network security management
system based on K-Means algorithm is designed. In terms of hardware, S9300 switch and RJ45 port
connector are designed. In terms of software, establish the function module of the network security
management system, establish the network management architecture according to the actual needs of the
system, and make the whole management process more reasonable. Optimize the density parameters of
network security management based on K-Means algorithm, analyze the initiative and enthusiasm of
network defense according to the network management architecture, and improve the network security
defense capability to the greatest extent, so as to achieve network security management. The system test
proves that the system has better performance and can be applied in real life.
1 INTRODUCTION
In the era of the Internet for All, Internet, big data,
cloud computing and other technologies have
gradually become essential technologies for people's
life, office and learning(Jiang C-Zhao D). All kinds
of financial services, such as payment and
settlement, financial investment, and life payment,
can be realized in the network environment. People
can buy, earn money, and pay fees online without
leaving home, which greatly improves the quality of
life of people. In addition, the network can also
provide people with travel route planning, hotel
restaurant reservation, food reservation and other
businesses. People only need to log on to relevant
websites to obtain the materials they need, which
plays an important role in optimizing people's travel
and life(Vajjha H-Al-Haija Q A). In terms of
functions, various network terminals can realize
various functions such as communication and
shopping; In terms of management, all kinds of
network terminals can carry out business
management, data management, data analysis, etc.,
which plays an important role in the development of
social economy.
At present, social enterprises have also begun to
develop network terminals to provide guarantee for
higher management of enterprises. The use of
network terminals can achieve efficient management
of enterprises and improve their operational
efficiency (Zhi and Li J). The threat of the network
is also great, similar to hackers, Trojan attacks,
network vulnerabilities and other security risks.
Hackers can steal deposit information in network
terminal software, damage the financial market, and
attack the servers of large government and enterprise
units through blackmail viruses, resulting in the
information of units being stolen, or unable to work
normally. The network can only be unlocked by
paying ransom(Wang M - Huang J). From this point
of view, network viruses pose a great threat to
society, spread widely and have a long hiding
period. It is urgent to design a security management
system to ensure the safe use of the network.
Therefore, this paper designs a network security
management system based on the function of
K-Means algorithm.
2 HARDWARE DESIGN
2.1 S9300 Switch
According to the actual requirements of the system,
the management system designed in this paper
divides the hardware structure of the system into
104
Liu, A.
Design of Network Security Management System Based on K-Means Algorithm.
DOI: 10.5220/0012275400003807
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 2nd International Seminar on Artificial Intelligence, Networking and Information Technology (ANIT 2023), pages 104-109
ISBN: 978-989-758-677-4
Proceedings Copyright © 2024 by SCITEPRESS – Science and Technology Publications, Lda.

sharing, connection and other modes, and the
sharing end is divided into device sharing and data
transmission sharing; The connection end is divided
into device connection and data connection(Chao H
C - Yuan T). In order to ensure the security of the
network security management system, this paper
designs the S9300 switch at the network
management end, which forms the system sharing
structure together with the out of band network
management server and the network security
firewall. As shown in Figure 1 below.
Figure 1: Functional Structure of S9300 Switch.
As shown in Figure 1, the S9300 switch designed
in this paper is a fully digital centralized control
switch that can run on mobile networks and public
networks(Zhang N - Li R). SPZ and SPT are the
control parts of S9300 switch, which control the
security environment for storage, analysis and
management of various network data. CPS, RPS,
IOS, MAS, etc. are the network data processing
subsystems of SPZ, which have the functions of
maintenance, processing, input and output, and play
an important role in improving the security
performance of the network management system.
2.2 Rj45 Port Connector
The RJ45 port connector designed in this paper is
mainly used to work with the S9300 switch. During
the operation of the network security management
system, the RJ45 port connector has the IEEE802.3
standard, and the maximum effective working
distance can reach 100m, thus ensuring the
effectiveness of the entire network management
environment (Zhang, 2021). Relevant parameters of
RJ45 port connector are shown in Table 1 below.
As shown in Table 1, this paper selects the
parameters described in the table to ensure the
storage and forwarding capability of network data
management (Melad T-Humayun M) under half
duplex, full duplex, self negotiation and other
working modes. When used together with the S9300
switch, it can change the switching mode of the
system to maximize the backplane bandwidth and
forwarding capability, thus ensuring the overall
performance security of the system.
Table 1: Parameters of RJ45 Port Connector.
classification
p
arameter
Fixed port
8 10/100/1000Base-T adaptive Ethernet
ports
Fixed Port
Properties
Connector type: RJ-45
Supports 10/100/1000Mbit/s transmission
rate
Supports half duplex, full duplex, and
self negotiation working modes
Supports MDI/MDI-X adaptation
Network cable
type
10/100Base-TX :
Category 3/4/5 twisted pair, supporting a
maximum transmission distance of 100m
1000Base-T :
Category 5 double line, supporting a
maximum transmission distance of 100m
pilot lamp
Per Port: Link/Act , Speed
Per device: Power
backplane
b
andwidth
16Gbps
Forwarding ability
11.9Mpps
Exchange mode
Store Forward Mode
MAC Address
Table
Supports address automatic learning and
aging (aging time is 5 minutes)
Maximum support for MAC (Medium
Access Control) addresses: 8
K
3 SOFTWARE DESIGN
3.1 Establish the Function Module of
Network Security Management
System
In the network security management function
module designed in this paper, each user is assigned
an account. The account enters the system through
the login terminal, connects to the network security
data acquisition system function, and realizes data
security management (Jiang L - Zhang Q). In this
paper, with the help of K-Means algorithm, relevant
business data is stored in the database, and Web
logic business processing services are used to
process, real-time transmission of network security
data. The functional architecture of the network
security management system is shown in Figure 2
below.
Design of Network Security Management System Based on K-Means Algorithm
105

Figure 2: Functional Architecture of Network Security
Management System.
As shown in Figure 2, this paper divides data
acquisition, data analysis and post-processing into
one functional module, which is mainly responsible
for resource allocation, load monitoring and other
functions. Then MySQL, Web service center, ESB
adaptation center, information transmission, etc. are
divided into a functional module, which is combined
with storage center, CPU, communication link,
auxiliary equipment and other functional modules to
form a virtualized and shared operation service
function. The ESB is used to monitor the data access
request of each network software, send the relevant
request to the big data center for analysis, and
realize the storage and processing of network
security video, image, voice and other data to ensure
network security to the greatest extent.
3.2 Optimize Network Security
Management Density Parameters
Based on K-Means Algorithm
For the network security management system, the
optimization of density parameters is effective in
network security defense and plays an important role
in network security management(Fang W, 2022).
This paper mainly uses the K-Means algorithm,
taking the network data object Xi as the center, and
the average distance of the data object nearest to Xi
is expressed as the density parameter of the network
data object Xi. In this paper, m-dist is used to
represent the optimized density parameter value. The
larger the m-dist value, the smaller the density, and
the more dispersed the data object Xi and the
adjacent data objects are. This paper gives the
network data sample Xi, which is expressed as:
{
}
() () ()
12
, ,...,
ii i
in
Xxx x= (1)
In formula (1),
i
X
give network data samples
for this paper; 𝑥

()
, 𝑥

()
, 𝑥

()
is of
i
X
discrete
data of. This paper uses the K-Means algorithm, and
the selection of k is determined according to the
actual network demand
i
X
the ordered attribute
distance from adjacent data objects is:
1
() ( )
1
(, )
n
k
k
ij
ki j u u
k
dist X X x x
=

=−



(2)
In formula (2), 𝑑𝑖𝑠𝑡

(𝑋

, 𝑋

) is a data
object𝑋

and adjacent data objects𝑋

ordered attribute
distance; 𝑥

()
、𝑥

()
by
i
X
and
j
X
sample data of.
On this basis, analyze
i
X
and
j
X
the unordered
attribute distance of is as follows:
() ( )
() ( )
() ( )
1
,, , ,
(, )
,,
k
ij
k
ij
uu uu
uu
ij
n
uu uu
mx n mx n
mx x
mx mx
=
=−

(3)
In equation (3),
() ( )
(, )
ij
uu
mx x
by
()i
u
x
and
()
j
u
x
the unordered attribute distance of;
()
,
i
uu
mx
is the
value of attribute u 𝑥

()
number of samples;
𝑚

, 𝑥

()
, 𝑛is the value of attribute u in the nth
sample cluster𝑥

()
number of samples;
()
,
j
uu
mx
is
the value of attribute u𝑥

()
number of samples;
()
,,
j
uu
mx n
is the value of attribute u in the nth
sample cluster 𝑥

()
number of samples.
Take 𝑑𝑖𝑠𝑡

(𝑋

, 𝑋

) and𝑚(𝑥

()
, 𝑥

()
) is expressed
as the optimized value of density parameter, and the
formula is as follows:
1
() ( ) () ( )
11
(, ) ( , )
nn
k
k
ij ij
ij u u u u
uun
mdistXX x x mx x
==+

−=−+



(4)
In equation (4), 𝑚−𝑑𝑖𝑠𝑡(𝑋

, 𝑋

) optimize
values for density parameters. In the process of
network security management, network data X
i
is
output, and the number of clusters is expressed as K.
The output result is K clusters of data X
i
. Select K
objects in the high-density area from the output
results as the cluster center, and execute the
K-Means algorithm,
(, )
ij
mdistXX−
the smaller
the value, the greater the density of network data, the
denser the data, and easier to implement the
K-Means algorithm to ensure network security.
4 SYSTEM TEST
In order to verify whether the system designed in
this paper has the security management
performance, this paper tests the above system.
ANIT 2023 - The International Seminar on Artificial Intelligence, Networking and Information Technology
106

Ensure that the system is in stable operation under
the environment where both hardware and software
are in normal operation. The traditional network
security management system and the management
system based on K-Means algorithm designed in this
paper are used to test the security management
performance of the system. The test preparation
process and the final test results are shown below.
4.1 Test Preparation
This paper analyzes the functional requirements of
the system, designs the network security
management process for the system functions, and
uses Java language to develop the system. The
system development tool is MyEclipse integrated
platform, and the Tomcat operating system is
Windows7. MySQL database is used for data
storage. The database table is shown in Table 2
below.
Table 2: MySQL database table.
attribute type explain Can be
empty
SLP_PhoneNo Char(11) System login port Yes
SLD_DevNo Char(10) System login device No
C_Pw
d
Varchar(10) cipher No
S_SimNo Char(10) Si
m
No
AAP_Adr Varchar(11) APN Access Point No
SLN_Name Varchar(10) System login name No
APT_Type Varchar(10) Access point type No
AT_IdentityType Char(4) Authentication Type No
NSMMT_MsgNo Varchar(10) Network Security Management
Message Types
No
NSMC_Content Varchar(50) Network security management
content
No
As shown in Figure 2, the database in this paper
follows the third normal form principle, and each
data depends on transmission, which can solve the
problem of network data redundancy, thus ensuring
the effective management of the system. After the
system environment is prepared, this paper
completes the hardware installation of S9300 switch
and RJ45 port connector according to the
instructions, and debugs the hardware. The S9300
switch is powered on after all lines are connected.
The voltage of each line of the switch is within the
range of 3.0V~5.0V, which can ensure the normal
use of the switch. After each line of RJ45 port
connector is connected, the green indicator of the
connector lights up to ensure the normal operation of
the connector. After the hardware debugging is
completed, this paper debugs the software, and the
debugging code is shown in Figure 3 below.
Figure 3: Software Debugging Part Code.
As shown in Figure 3, in the process of software
debugging, vulnerabilities were found in the
software program, and K-Means algorithm was used
to timely correct the code, make up for the
vulnerabilities, and improve the running reliability
of the software. After the debugging of hardware and
software is completed, the system is in normal
operation state. In this paper, enter the correct login
name and password. After logging into the system,
the interface shown in Figure 4 appears.
Figure 4: Network Security Analysis Interface of the
System.
As shown in Figure 4, after clicking "Statistics"
for network activity protocol statistics, network
security management starts to run. Analyze the
forwarding data, error data, time stamp and other
data to determine the network environment, and
extract the unreasonable and wrong network data to
ensure the normal operation of the system.
4.2 Test Results
Under the above test conditions, this paper randomly
selects 8 types of system operation information to
analyze the system operation stability and operation
standardization. Among them, the time spent in
Design of Network Security Management System Based on K-Means Algorithm
107

system testing, the throughput of relevant operation
information owned by the system during execution,
the throughput of network security management
system information, the number of hits of network
security management system execution information,
the number of hits of network security management
system, and other operation information are the
information to analyze the stability of system
operation; Enter the user name admin and password
admin, enter the non-existent user name abc, enter
the password 123abb589, enter the IP address, and
obtain the operation information such as data
packets under the IP address, which is the normative
information of the analysis system operation. The
traditional network security management system was
used to test with the network security management
system based on K-Means algorithm designed in this
paper. The test results are shown in Table 3 below.
Table 3: Test Results.
System operation information Traditional management
system performance
The performance of the
management system
designed in this article
Time spent on system testing 78s 32s
The throughput of relevant
operation information possessed
b
y the system during execution
106M 43M
Network Security Management
System Information Throughput
0.82M/s 1.34M/s
Number of clicks on network
security management system
execution information
2045 times 6025 times
Number of clicks in the network
security management system
96.3 times 224.7 times
Enter username admin and
password admin
After two failed login
attempts, successfully
log in and display the
system homepage
content
Successfully logged in
and displayed the
system homepage
content
Enter a non-existent username
abc and a password of
123abb589
Prompt for password
error
Prompt that the
username does not
exist or the password is
incorrect
Enter an IP address to obtain
data packets under that IP
address
After prompting for an
IP address that does not
exist 3 times, display
information about the
data packets related to
the input IP address
Display information
about input IP address
related packets
As shown in Table 3, this paper randomly selects
eight types of system operation information,
including the time spent in system testing, the
throughput of relevant operation information held
during system execution, the throughput of network
security management system information, the
number of hits of network security management
system execution information, the number of hits of
network security management system existence, the
user name admin and password admin enter the
non-existent user name abc, enter the password
123abb589, enter the IP address, and obtain the
operation information such as data packets under the
IP address, which is the key information to reflect
the system operation performance. It can be seen
from the table that after using the traditional network
security management system, the system
performance is poor, the system operation interface
is non-standard, and the system cannot be operated
accurately, affecting the use of the system. After
using the network security management system
based on K-Means algorithm designed in this paper,
the system performance is better, the system running
interface is smoother, and the system can be
operated accurately to ensure the effective
management of the network.
5 CONCLUSION
In recent years, Internet technology has developed
rapidly, and network terminals are used more
frequently, which not only changes people's
lifestyle, but also provides a more convenient living
environment for people. The network business in the
enterprise is applied on the network terminal. Only
personnel inside the enterprise can change the
network data in the terminal to ensure the safe
storage of enterprise data. In addition, network
business greatly saves time costs and improves the
efficiency of managers. However, the emergence of
the network has not only brought convenience, but
also brought hidden information for people. The
problem of malicious attacks on the network
continues to appear, trojans, network vulnerabilities
and other problems, making people fall into a
network crisis. Therefore, this paper uses K-Means
algorithm to design a network security management
system. Combine the hardware and software of the
system with the actual functional requirements of the
system to truly improve the security of network
management.
REFERENCES
Jiang C.Network Security and Ideological Security Based
on Wireless Communication and Big Data Analysis[J].
Wireless Communications and Mobile Computing,
2022, 2022(3):1-6.
Liang W , Xie S , Cai J , et al. Deep Neural Network
Security Collaborative Filtering Scheme for Service
Recommendation in Intelligent Cyber-Physical
Systems(J]. IEEE Internet of Things Journal, 2021,
PP(99):1-1.
Zhao D , Song H , Li H . Fuzzy integrated rough set theory
situation feature extraction of network security[J].
Journal of Intelligent and Fuzzy Systems, 2021,
40(1):1-12.
ANIT 2023 - The International Seminar on Artificial Intelligence, Networking and Information Technology
108

Vajjha H , Sushma P . Techniques and Limitations in
Securing the Log Files to Enhance Network Security
and Monitoring[J]. Solid State Technology, 2021,
64(2):1-8.
Tang C , Zhang J , Wang S , et al. Application and
Implementation of Big Data Visualization Technology
in Network Security System[J]. Journal of Physics:
Conference Series, 2021, 1955(1):012002 (6pp).
Al-Haija Q A , Ishtaiwi A . Multi-Class Classification of
Firewall Log Files Using Shallow Neural Network for
Network Security Applications[J]. Advances in
Intelligent Systems and Computing, 2021,
1370(1):1-15.
Zhi W W , Zhou X X , Yang L . Application of Fuzzy
Comprehensive Method and Analytic Hierarchy
Process in the Evaluation of Network Security Level
Protection Research[J]. Journal of Physics: Conference
Series, 2021, 1820(1):012187 (8pp).
Li S . Development Trend of Computer Network Security
Technology Based on the Big Data Era[J]. Journal of
Physics Conference Series, 2021, 1744(4):042223.
Li J , Yan L , Wang J , et al. Research on Network Security
Risk Assessment Method Based on Improved AHP[J].
Journal of Physics: Conference Series, 2021,
1828(1):012115 (14pp).
Wang M , Zhang B . Analysis of Internet of Things
Computer Network Security and Remote Control
Technology[J]. IOP Conference Series: Earth and
Environmental Science, 2021, 634(1):012035 (6pp).
Wang D , Fan H , Li G , et al. Research on Real-time
Crawling and Marking Technology of Sensitive Access
SQL Statements Based on Information Network
Security Isolation Device[J]. Journal of Physics
Conference Series, 2021, 1948(1):012059.
Huang J , Li J , Wang J , et al. Information Hiding
Technology under the Background of Power System
Network Security[J]. Journal of Physics: Conference
Series, 2021, 1852(2):022064 (7pp).
Chao H C , Wu H T , Tseng F H.AIS Meets IoT: A
Network Security Mechanism of Sustainable Marine
Resource Based on Edge Computing[J]. Sustainability,
2021, 13(6):3048.
Einy S , Oz C , Navaei Y D.The Anomaly- and
Signature-Based IDS for Network Security Using
Hybrid Inference Systems[J]. Mathematical Problems
in Engineering, 2021, 2021(9):1-10.
Yuan T , Niu Y , Lv L.Research on Optimization Strategy
of Computer Network Security Technology under the
Background of Big Data Era[J]. Journal of Physics:
Conference Series, 2021, 1744(3):032237 (4pp).
Zhang N . Research on the Application of Data Encryption
Technology Based on Network Security Maintenance
in Computer Network Security[J]. Journal of Physics
Conference Series, 2021, 1744(2):022060.
Ma J , Yu G , Ke H , et al. Network Security Management
and Protection of Industrial Internet Equipment[J].
Chinese Journal of Engineering Science, 2021,
23(2):81.
Li R , Li F , Wu C , et al. Research on Vehicle Network
Security Situation Prediction Based on Improved
CLPSO-RBF[J]. Journal of Physics: Conference Series,
2021, 1757(1):012148 (8pp).
Zhang H , Kang C , Xiao Y.Research on Network Security
Situation Awareness Based on the LSTM-DT
Model[J].Sensors, 2021, 21(14):4788.
Melad T.Data Network Security Over The Horizon
Radar[J]. International Journal of Scientific and
Research Publications (IJSRP), 2021, 11(7):730-746.
Humayun M , Hamid B , Jhanjhi N Z , et al. 5G Network
Security Issues, Challenges, Opportunities and Future
Directions: A Survey[J]. Journal of Physics Conference
Series, 2021, 1979(1):012037.
Jiang L.On the Relationship between Computer Firewall
Technology and Network Security[J]. Journal of
Physics Conference Series, 2021, 1744(3):032137.
Ni Z , Zhao F.Research and Implementation of Network
Security Management Based on Virtualization
Technology[J]. Journal of Physics Conference Series,
2021, 1802(4):042070.
Zhang Q , Ma D. Discussion on Network Security of
Information Retrieval Technology in Network
Communication[J]. Journal of Physics: Conference
Series, 2021, 1744(3):032200 (6pp).
Fang W , Zhao R , Zhu D.Study on Sparrow Search
Algorithm Based on K-Means Clustering[J]. Computer
Simulation, 2022, 39(9): 403-409.
Design of Network Security Management System Based on K-Means Algorithm
109