43
1 INTRODUCTION
The COVID-19 recession caused an increase in
cyberattacks on information systems around the
world. Given that shipping transports 90% of global
trade, the maritime industry and maritime transport
infrastructure are very attractive targets for
cybercriminals. For example, during hybrid wars
hackers can disrupt the delivery of supplies to cause
massive damage to the country's economy. The Israeli
company NavalDome reported a 400% increase in
attempted cyberattacks on information systems in
2020. In general, the damage caused by cyberattacks
cost the global maritime industry about $200 billion.
A global problem is the inability to update ship
security systems when on a voyage or in remote ports,
or in the roadstead. Shipowners have to wait for a
ship to complete her voyage or call at a port that has
the necessary information resources and equipment
[20].
Due to social constraints, recession, and abundance
of malicious software (ransomware, phishing attacks,
social engineering techniques, etc.) shipping
companies are not able to sufficiently protect
themselves. This forces suppliers, equipment
manufacturers, and IT professionals to connect
autonomous systems to the Internet to ensure timely
Ensuring Cyber Resilience of Ship Information Systems
O. Onishchenko
1
, K. Shumilova
1
, S. Volyanskyy
2
, Y. Volyanskaya
2
& Y. Volianskyi
2
1
National University Odessa Maritime Academy, Ukraine
2
Admiral Makarov National University of Shipbuilding, Ukraine
ABSTRACT: The Covid-19 pandemic brought a problem related to the inability to timely update security
systems on ships during a voyage and the need to encrypt data stored in ship information systems (SISs) and
shipping company information systems. The analysis of new types of worldwide cyberattacks showed that they
were set off by an increase in the use of remote-controlled autonomous technologies and a spread of telework. It
was proved that on ships: a) there are no cybersecurity specialists; b) there is no means of identification of cyber
threats and no response plans; c) there is a lack of cybersecurity training for crews; d) encryption of confidential
ship data is barely used; e) crew is a vulnerability factor in ship's security. The analysis of cyber incidents
allowed us to develop a basic response plan to protect ship control systems. It was demonstrated that the basic
plan can be continuously updated and improved in accordance with: a) the real state of ship systems; b) the
results of performance analysis of crew actions; c) the emergence of new types of cyberattacks. To improve the
security of confidential data in ship information networks theoretical framework for the development of
encrypted data search engines with the identification of “dangerous” keywords for the ship information
systems (SISs) was proposed. A data exchange protocol, basic requirements for SISs, and an algorithm for
detecting “dangerous” keywords in messages were developed. A test search engine on encrypted data was
presented, and the main components of the system were highlighted. The functionality of the system was
experimentally proved, and the accuracy and speed of search on encrypted data were determined.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 16
Number 1
March 2022
DOI: 10.12716/1001.16.01.04
44
maintenance, storage, and distribution of information.
Such autonomous systems and processes increase the
vulnerability of maritime information networks
leading to the possible emergence of new types of
cyberattacks.
Operators of coastal and marine networks and
systems are usually convinced that a traditional
antivirus system protects them against cyberattacks
and blame any anomaly on the need to reboot servers,
a system error or a system failure. However,
individual systems in operational technology (OT)
network are insecure because firewall and software
only protect information technology (IT) resources.
Therefore, individual OT endpoints, critical systems,
and components may be sensitive or obsolete, lacking
security updates, which increases vulnerability to
cyberattacks. “Computers, servers, laptops, tablets,
mobile phones and other devices are considered to be
endpoints. The lack of reliable endpoint protection
made it possible to launch such attacks as Petya,
WannaCry, and Bad Rabbit” [9].
Any part of the ship traffic management system, as
well as the cargo handling and security system, can
become vulnerable. Protecting the entire network
from attacks will not solve the vulnerability problem.
You need to protect each information system, encrypt
its data, and adopt advanced user authentication
methods. Unless there is an understanding of the
scope and severity of relevant cyber risks, the ship-
shore system will be deadlocked “no vision– no
actions”. The OT network has no “dashboard” to
allow operators to see the status of all systems
connected. With vision comes the ability to take
adequate actions and respond to cyber threats. Even
some baseline monitoring and response plan will
make it considerably harder to carry out a ship
shore” cyberattack, and the resilience of the ship's
information system needs to ensure the reliability of
data storage (commercial, logistics, controlling
technical systems, etc.) on ship servers and local
devices.
It should be recognised that a ship information
system (SIS) is a complex technical system and its
behaviour is described by nonlinear interconnections
and complex interactions with the environment [2, 7,
10]. SIS has specific features: nonlinearity,
heterogeneity, uncertainty, stochasticity, and cyclicity
[11]. The variety of types of ships has led to the fact
that the structures of modern SISs differ significantly,
and developers face a number of serious problems
related, in particular, to conducting a qualitative and
quantitative analysis of systems efficiency in the initial
stages of design.
When synthesizing modern SISs, it is necessary to
take into account the following factors and
parameters: a) complexity a holistic approach to
automation of technological processes on a ship; b)
efficiency speed of processing and availability of SIS
data; c) flexibility the ability to quickly change the
configuration or functional sets of SISs depending on
external environment; d) distribution a multilayer
structure and hierarchy of SIS servers; e)
interconnection with other networks the ability to
import and export data arrays in widely accepted data
exchange formats; f) data openness.
The last point is of particular importance and
forms a serious contradiction the need to increase
the degree of openness for external users and the need
to protect your information. A modern SIS or
shipping company system must have mechanisms for
sharing its data over the Internet price lists, a list of
services, ads, inside information, etc. It is clear that
developers do not make all data publicly available
and therefore special emphasis should be placed on
protection SISs to prevent unauthorized access to
business data, technical services, control and
identification systems, and ship devices [6], for proper
organization of information access levels.
The relevance of the research topic is connected
with the global crisis and social distancing measures
that prevent IT specialists from being mobile in the
maritime sector and upgrading and maintaining
critical ship OT systems promptly. Such a situation
makes operators neglect security protocols and
therefore ship's control systems and information
networks become vulnerable to hacker attacks.
2 LITERATURE REVIEW AND PROBLEM
STATEMENT
The research [18] by the maritime cybersecurity
company CyberOwl presented at the CyberSecure at
Sea conference showed the results of a survey of 120
IT specialists on cargo shipping. It has been
demonstrated that most specialists lack understanding
not only of the problems of protecting their ship's
networks and devices but also of their overall
structure. Some of them have poor central visibility.
Some identified a lot more opportunities for
connecting “shadow” IT on board the ship. CyberOwl
reports that virtual blindness and lack of data
protection became the current shipping reality.
[22] states that long-term cybersecurity projects are
difficult to implement. They are based on
comprehensive risk assessment, change of network
architecture to improve segmentation, controls
updating and risk management analysis. The issues of
assessment of cyber risks and cyberattacks
consequences on each ship remain unresolved.
As reported on [21], many people still are not able
to detect even the simplest phishing emails used by
hackers to steal personal and corporate information
(via email, messages in social networks, fake websites,
etc.). Even charging a smartphone with a USB port via
an ECDIS terminal (ECDIS, Electronic Chart Display
and Information System) can grant hackers access to
the ship's information systems and lead to data
leakage.
The reason may be not only the lack of
cybersecurity specialists and an up-to-date cyber
incident response plan on board but also exposure of
confidential information, the lack of even primitive
protection. The mentioned challenges may be
overcome through risks identification on each
particular ship, development of appropriate
cyberattack response plans for the crew, and data
encryption. This approach is applied in [18], which
defines the categories of cybersecurity procedures and
45
necessary actions that can be used to train ship
personnel and prevent cyber threats.
A webinar [1], hosted by the Aspen Institute
(USA), informed that since the COVID-19 pandemic
started, the number of cybercrimes reported to the
Crime Complaint Center of the Federal Bureau of
Investigation (FBI) has roughly quadrupled. The
biggest cyberattacks were carried out by hostile
foreign entities or intelligence agencies. The source [4]
shows that the main problem is that many people
work remotely, so viruses encountered by employees
easily spread to their personal devices. The most
common websites attacks [12] are malicious code
injection (SQL Injection), Path Traversal and Cross-
Site Scripting (XSS). This is what leads to data leakage
from local networks and individual devices.
The source [17] describes a new type of DDOS
attack, which became the biggest in history and
caused a 30-minute shutdown of 15% of the global
Internet and a number of backbone providers.
This suggests that it would be appropriate to
conduct a study devoted to a) the development of a
response plan to modern cyber incidents for ship
management information systems; b) the
improvement of methods for storing and encrypting
confidential data on SIS.
3 THE AIM AND OBJECTIVES OF THE STUDY
The study aims to: a) identify areas of vulnerability of
maritime information systems to cyber threats, taking
into account the likelihood of cyberattacks and their
consequences for navigation processes; b) determine
the basic processes for monitoring the cyber resilience
of information systems and develop a basic response
plan for the ship's crew; c) improve algorithms for SIS
data encryption.
Therefore, it is necessary to ensure the SIS cyber
resilience to create conditions for the functioning of
the system before, during and after a cyberattack,
including information attacks. To solve the
abovementioned problems, it is necessary to conduct a
meta-analysis that will allow using original research
data from international IT companies, summarize the
results devoted to the cybersecurity problem and
protection of systems in OT network.
To achieve this, the following objectives were set:
carry out a meta-analysis of the maritime
information cyberspace context;
identify prerequisites for the development of a
crew`s response plan to information risks and
prevention of critical ship safety management
systems (SMS) failure;
improve the SIS data encryption system.
4 THE STUDY MATERIALS AND METHODS
The study uses the following methods: 1) the method
of passive monitoring and risk identification in the
development of a crew’s response plan to information
risks; 2) homomorphic data encryption, probabilistic
Monte Carlo algorithms in a cryptographic
protection system for data stored on SIS servers
(described in section 5).
4.1 Examination of the vulnerability of maritime
information systems from the perspective of
cybersecurity of shipping processes
It should be recalled that cybersecurity aims to protect
information systems, networks, and programs from
cyberattacks. Based on the analysis of data from
cybersecurity experts in the maritime industry
(Admiral Makarov National University of
Shipbuilding, National University “Odessa Maritime
Academy”, Positive Technologies company), we will
identify critical information systems:
AIS Automatic Identification System;
ECDIS Electronic Chart Display and Information
System;
VDR Voyage Data Recorder;
TOS Terminal Operating System;
CTS Container Tracking System;
EPIRB Emergency Position Indicating Radio
Beacon;
GNSS Global Navigation Satellite System;
GPS Global Positioning System.
Literature and Internet sources, maritime practice
discovered that the most vulnerable ship systems are:
bridge systems; cargo handling systems; engine,
machine and power control systems, data storage and
processing systems. Access control systems, passenger
service and management systems, public Internet
networks, administrative systems and networks, and
communication systems are also at risk.
It is important to emphasize that any ship as well
as all ship navigation equipment and ECDIS systems
can fall a victim to a cyberattack, even at sea.
4.2 Meta-analysis of actual cyber incidents and
consequences of cyberattacks for monitoring and
analysis
The summary of cyberattacks notifications for 2017-
2021 [8, 15] found that companies do not provide a
detailed report on what happened. According to
Reuters, little information about successful attacks on
ships is publicly disclosed: business owners often
cover them up for fear of damaging the image,
receiving claims from customers and insurance
companies, investigations being initiated. Between
2017 and 2021 Barcelona and San Diego ports were
targeted by cyberattacks. Then the same happened to
the Australian shipbuilder Austal, the three largest
container carriers Maersk, COSCO and MSC, as well
as the Shahid Rajaee port in Iran.
Maersk, June 2017 a major cyberattack on APM
Terminals system caused by the NotPetya malware.
The attack disrupted the operation of ports and
terminals, which sustained losses of approximately
$300 million. This system handled more than 100 000
containers per day and was completely paralyzed,
which led to failures in the container turnover
schedule and huge losses. Maersk container ships
stopped at sea and 76 port terminals of the company
all over the globe were also stopped.
46
COSCO, July 2018 an attack on the company’s
digital assets caused the shutdown of email and
telephone systems, connections with other regions
and took down half of the shipowner’s US network.
MSC, April 2020 hit by the Ruik ransomware that
brought down the MSC website, caused a partial
shutdown of servers at the company’s headquarters in
Geneva; websites MSC.com and MyMSC became
unavailable due to data centre closure. Shahid Rajaee
port, May 9, 2020 a cyberattack on the port terminal
hacked the key OT systems. Shipping stopped
computers that regulate ship traffic, trucks and goods
crashed.
According to the research by the cybersecurity
centre Positive Research [19], the most common
attacks on websites are “malicious code embedding”
SQL Injection [12], “breaking out of a directory”
Path Traversal. The common virus Cross-Site
Scripting allows accessing the “administration panel”.
5 PREREQUISITES FOR ESTABLISHING BASIC
CYBER RESILIENCE MONITORING OF SHIP
INFORMATION SYSTEMS
5.1 Analysis of modern-day cyberattacks to define cyber
threat response processes
Various maritime organizations are active in
cybersecurity regulation of the global maritime sector,
including IMO (International Maritime Organization),
ISO (International Organization for Standardization),
IACS (International Association of Classification
Societies). But most internet technology professionals
admit that they are not aware of the design of ship
networks and how they are connected to distributed
operational technology networks, to business
networks, what vulnerabilities they cause, and how
cyber threats spread across them (Figure 1.).
As shown in Figure 1., 75% of specialists have no
understanding of their ship networks, 38% discovered
possibilities to connect “shadow IT”, and 36% lack
central visibility into networks and connected devices.
Based on the study of cyberattack types published in
international materials, as shown in Figure 2., we will
identify the most dangerous ones for the global
maritime sector.
- there is no understanding of their ship networks and devices;
- more possibilities of connection of "shadow" IT are revealed;
- in sufficient central visibility.
Figure 1. Visibility of ship networks and devices for IT
professionals, 2020
The common types of cyberattacks shown in
Figure 2. Indicate that the highest proportion of
cyberattacks includes: “malicious code injection”
(SQL Injection) 27%, “breaking out of a directory”
(Path Traversal) 17%, “cross-site scripting” (XSS)
14%.
Such modern viruses as Bruteforce, Petya,
WannaCry, Bad Rabbit are known for causing
millions in damages for shipping companies. In 2017,
the WannaCry ransomware caused 1 billion dollars in
damage and managed to infect 500 000 computers in
150 countries around the world.
Figure 2. Widespread global cyberattacks on web
applications
The Petya virus encrypted database files and data
relevant for operating system startup and then
demanded a bitcoin ransom. It affected government
agencies and companies in Europe, the USA,
Australia, Ukraine, India, Russia, and China.
5.2 Identification of processes for the development of cyber
resilience monitoring and response plan
Taking into account the abovementioned
consequences of a cyber security breach, we will
highlight the main processes necessary for designing a
basic response plan as shown in Table 2.
The developed processes will make it possible to
implement a response plan set up to address a certain
type of cyberattack, estimate the time required to
restore the ship’s control system, and use analysis to
demonstrate the efficiency of these processes in
reducing cyber risks.
5.3 Secure data storage and processing on ships
Encryption a well-known method used to protect
the confidentiality of information. Formerly, all
information stored in the SIS was to be encrypted.
Modern cryptography provides an extensive set of
encryption schemes. At the same time, to ensure the
highest SIS security degree all data has to be
processed in encrypted form. To solve tasks like these,
homomorphic encryption is used. It allows not only
performing encryption and decryption of data but
also analyzing it without disclosing the contents to
anyone [13].
Modern mail servers, such as IMAP servers, file
servers, and other storage servers, usually have to be
totally reliable. Users need to be sure that their data is
not disclosed without their permission, which poses
undesirable security and privacy risks. The
fundamental problem is that transferring calculations
to a data warehouse appears to be very time-
consuming, especially if data is encrypted, and many
computation tasks over encrypted data previously
had no practical solution.
47
Table 2. Processes of building a basic plan for responding to cyber threats of the ship SISs
__________________________________________________________________________________________________
Processes Necessary actions
__________________________________________________________________________________________________
1. Identification Description of the threat, incident, including automatic collection and aggregation of data from a set of
of the incident, monitoring sources. It allows you to quickly identify system cross-links and analyze information, identify
cyber threat, and rank risks, and visualize potential losses from cyber attacks.
vulnerability
2. Assessment Quantitative assessment of incidents and threats. Creating a registry of cyber incidents and vulnerabilities.
of the degree Identification of areas of vulnerability of systems. Measurement of system failures and detection of abuse
of danger of usage policy. This will help identify incorrect system configuration or suspicious system behavior.
Therefore, this process visualizes a vulnerability to deploying a new server or adding a new link, mobile
application, or web service. It will also give an idea of the quality of staff training for cyberattacks.
3. Determining Priority actions: detection of a cyber threat or cyber incident, minimization of the probability of their
the level of occurrence, restoration of the OT system. This will reveal the interdependence between critical systems
and
protection of control open access ports. Protection against the next generation of threats: update of anti-virus protection,
critical systems the presence of high-level firewalls application-level proxy servers (firewall). Regular updating of system
data with the introduction of new cybersecurity products. Mandatory certification of devices and
equipment used by the pilot, checking the tablets for the absence of unnecessary (undeclared) electronic
implants and prompt review (review) of the program code by the control service.
4. Isolation. Isolation of “infected” equipment and immediate notification of authorities, shore services and crew about
Documentation the incident. Ensuring unimpeded access to officials and competent authorities to restore the system as
soon as possible. Attracting the help of experienced IT consultants. Using a backup database outside the
internal platform. Documenting the incident: recording the time of its occurrence to detect infected systems
and data leakage. Involvement of forensic scientists. Surveys of persons involved in the cyber incident.
5. Analysis of Early detection of cyberattacks to prevent burglary and prevent failure of critical ship management
cyber resilience systems. Will protect marine systems by: identifying areas of vulnerability; risk ranking; identification and
authentication of all users; visualization of potential losses from possible cyberattacks; determining the
unity of the pilot service, shore services and AIS.
__________________________________________________________________________________________________
One solution to this problem in SISs includes
receiving files from a server and their decryption.
Another solution is to leave all data on the server and
develop a search algorithm for encrypted data using
methods for indexing each element or sequential
scanning without indexing. The disadvantage of
indexes is that their storing and updating requires a
significant amount of time. Also, this method is not
adequate for read-only data. Thus, the best option is
to perform a sequential scan without element
indexing. This method appeared in modern
cryptography for searching on encrypted data and can
be modified for SISs (Figure 3.).
Server
Data store
Search on encrypted data
[[𝑥]] = ([[𝑥
1
]], …, [[𝑥
𝑛
]])
User
Data preparation
p
k
, 𝑥 = (𝑥
1
..., 𝑥
𝑛
)
Decryption
s
k
Encryption and download
[[𝑥]] = ([[𝑥
1
]], …, [[𝑥
𝑛
]])
Returning anencrypted response [[𝑥
𝑖
]]
Encrypting and downloading
the request [[q]]
Search
s
k,
p
k
, q
Figure 3. Search scheme on encrypted data
For searchable encryption systems it is crucial to
demonstrate their ability to preserve the
confidentiality of user data and prevent information
leakage from SISs. Therefore, the resilience of
encryption systems to possible internal or external
attacks on an unreliable server as to be tested. The
server should not be able to learn anything about the
original data from the encrypted text or search
process. Data breach statistics [16] are shown in Figure
4.
We would highlight the main security features [16]
each searchable encryption system must support: a)
controlled searching unauthorized users should not
be able to perform server search [3, 5]; b) hidden
query a technique that hides an unencrypted query
from an untrusted server; c) query isolation in the
search process the server should know nothing, except
the search results. If there is a match between the
query and the index, the server can locate the related
documents and return them to the data user.
Number of
information leaks
Year
Figure 4. Annual number of information leaks in the world
An algorithm for searching over encrypted data on
the SIS server is introduced (Figure 5.).
This algorithm facilitates the checking for the
occurrence of certain (for example, “dangerous”)
keywords on encrypted data. For example, this makes
it possible to make up a psychological portrait of a
user of the developed system and determine (predict)
their further actions without revealing the user’s
identity. Thus, neither client (data owner) nor their
data will be publicly available and the system will be
able to clearly identify whether a user is a ship
specialist or a hacker.
To implement the developed message passing
system with search over encrypted data, Linux Mint
19.3 OS on Virtual Box was used (configuration 4
cores for the processor, 8 GB of RAM, 60 GB for hard
disk space). Clion which supports Cmake projects was
used as a development environment. The Helib
library was chosen as a library supporting
homomorphic encryption.
48
To test the system, the possibility of obtaining
public and private keys from client data on SIS server
side was checked. The process is shown in Figure 6.
Testing confirmed that a private key obtained on the
server side cannot decrypt data encrypted on the
client side.
Figure 5. Flowchart of the algorithm for checking messages
for the presence of “dangerous” words
Figure 6. Testing keys during exchange between the client
and the ISMS server
The next stage of the testing process was to check
the speed of server search for suspicious keywords in
a client message. This test is described by the
algorithm shown in Figure 7.
A graph in Fig. 8 shows how search speed changes
with increasing of data volumes. Theoretically, the
search graph should be represented by a logarithmic
curve, since it is based on the Binary Raffle Protocol
algorithm.
However, given that besides search the results are
subject to manipulations, the graph deviates from the
theoretical line (deviation does not exceed 5%).
Figure 7. Testing the speed of the search algorithm
Figure 8. Algorithm running time depending on the number
of words
Figure 9 shows the plot of the number of
“dangerous” keywords found against the number of
such keywords available in data. It can be observed
that the error is not permanent but probabilistic. It
should also be noted that the user's writing style
affects the search of “dangerous” keywords.
Another result of testing was that in some cases
certain keywords were not found. The reason is that
the homomorphic search scheme uses Monte Carlo
probabilistic algorithms, which perform a search
based on a set error probability. Here the error value
was taken as ε = 2-2, which is the biggest tolerable
error for this algorithm.
- found words; - the number of available words in the text.
Figure 9. Number of words found and available
The time spent on “dangerous” keywords
searching increases with the decrease of error
probability. This is not critical when dealing with
small data volumes, but as the volume increases, the
search speed decreases. However, this does not
reduce the practical value of the introduced
49
algorithms for designing dataware and software for
storing encrypted data in SISs.
6 DISCUSSION OF RESULTS
The obtained results demonstrate the need to identify
[6] and assess risks and identify vulnerable areas of
ship systems in terms of information security,
encryption of confidential information stored in SISs.
The specific feature of the introduced response
plan is that it is easy for ship personnel to understand
compared to multi-process and general
comprehensive cybersecurity measures. The flexibility
of the included procedures allows updating and
improving documentation, taking into account new
cyber incidents, changes in the state of, for example,
ship's power control systems [14], etc.
Further challenges in the development and
practical application of this study are related to the
lack of public access to information about cyberattacks
on each ship, lack of cybersecurity awareness among
seafarers, the need to encrypt data stored on ship or
company servers, and subsequent erroneous
decisions, information leakage. Thus, it creates a
problem of incident identification and reduces the
quality of systems state monitoring. The difficulty of
incident identification and monitoring may arise from
the inability to control all connected devices on the
ship, risk of ship personnel failure to follow existing
security procedures, and lack of response plans that
combine shore and sea cyber ecosystems to ensure
effective system recovery.
7 CONCLUSIONS
4. The results of the study are based on the analysis of
the most dangerous cyberattacks. The survey of
international IT specialists showed the lack of
awareness among modern companies about
protection against cyber threats, dangers
associated with leakage of confidential
information. Maritime industry practitioners have
little knowledge of the complexity of ship and
shore information networks. Unintentional actions
of the crew were determined to be the most
dangerous weakness since it is a seafarer who
allows a virus into equipment or clicks on
malicious links. The basic response plan
introduced in the study allows identifying and
assessing risks and can be continuously updated
depending on the identified areas of system
vulnerabilities.
5. Based on the analysis of existing search engines on
encrypted data, the criteria for the development of
secure search algorithms for SISs with the use of
homomorphic encryption were highlighted. A test
search engine on encrypted data was developed,
and its main components were identified. Software
implementations were made in CLion
development environment. The search engine for
“dangerous” keywords was tested. The study
introduced the solution for further improvement
of the algorithm for searching dangerous”
keywords over encrypted data by examining and
drawing up “dangerous” keywords rules in order
to predict their occurrence in the text, thus
allowing taking into account the writing style of
each SIS user, reducing the time spent on searching
and processing encrypted data. The security of
distributed SIS accesskeys was experimentally
proved, and the accuracy and speed of searching
over encrypted data were determined. The error of
the experimental speed curve deviation from the
theoretical curve does not exceed 5%.
REFERENCES
1. Apr 19, D.H.K.•, 2020: FBI: Covid-19 Cyberattacks Spike
400% in Pandemic,
https://www.msspalert.com/cybersecurity-news/fbi-
covid-19-cyberattacks-spike-400-in-pandemic/, last
accessed 2022/03/25.
2. Bar-Yam, Y.: General Features of Complex Systems.
UNESCO Encyclopedia of Life Support Systems. (2002).
3. Ding, M. et al.: An efficient Public Key Encryption with
Conjunctive Keyword Search scheme based on pairings.
In: 2012 3rd IEEE International Conference on Network
Infrastructure and Digital Content. pp. 526530 (2012).
https://doi.org/10.1109/ICNIDC.2012.6418809.
4. Ferrill, T.: 10 security tools all remote employees should
have, https://www.csoonline.com/article/3625882/10-
security-tools-all-remote-employees-should-have.html,
last accessed 2022/03/25.
5. Goh, E.-J.: Secure Indexes. (2003).
6. Golikov, V.A. et al.: A simple technique for identifying
vessel model parameters. IOP Conference Series: Earth
and Environmental Science. 172, 012010 (2018).
https://doi.org/10.1088/1755-1315/172/1/012010.
7. Kołowrocki, K., Soszyńska-Budny, J.: Reliability and
Safety of Complex Technical Systems and Processes.
Springer London, London (2011).
https://doi.org/10.1007/978-0-85729-694-8_8.
8. Kulesh, S.: NCCC NSDC recorded a new type of DDOS
attacks using hacked smart home devices?,
https://itc.ua/news/nkczk-snbo-zafiksiroval-novyj-tip-
ddos-atak-s-ispolzovaniem-vzlomannyh-ustrojstv-
umnogo-doma/, last accessed 2022/03/25.
9. Shumilova, K., Onishchenko, O.: Action planning in
comprehensive shipping risk identification. The
scientific heritage. The scientific heritage. 49, 1, 4046
(2020).
10. Tsvetkov, V.Y.: Complex technical systems. International
Journal of Applied and Basic Research. 10, 4, 670 (2016).
11. Tugolukov, E.N. et al.: Design of complex systems.
Tambov: ТGТU (2008).
12. VanMSFT: SQL code injection,
https://docs.microsoft.com/ru-ru/sql/relational-
databases/security/sql-injection, last accessed 2022/03/25.
13. Volianskyi, Y.: Development of practical implementation
of the use of schemes of homomorphic search. Slovak
international scientific journal. 48, 1, 712 (2021).
14. Volyanskaya, Y. et al.: Determining energy-efficient
operation modes of the propulsion electrical motor of an
autonomous swimming apparatus. EEJET. 6, 8 (90), 11
16 (2017). https://doi.org/10.15587/1729-
4061.2017.118984.
15. Analytical review of the protocols of the Internet of
Things, https://www.tssonline.ru, last accessed
2022/03/25.
16. Analytics of the information security industry 2021,
https://www.infowatch.ru/analytics/daydzhesty-i-
obzory, last accessed 2022/03/25.
17. Attacks on web applications: results of 2018,
https://www.ptsecurity.com/ru-
50
ru/research/analytics/web-application-attacks-2019/, last
accessed 2022/03/25.
18. Gaining visibility of your onboard systems: you can’t
secure something you can’t see,
https://thedigitalship.com/news/maritime-satellite-
communications/item/6673-gaining-visibility-of-your-
onboard-systems-you-can-t-secure-something-you-can-t-
see3, last accessed 2022/03/25.
19. Positive Research Center,
https://www.securitylab.ru/lab/, last accessed
2022/03/25.
20. Relevant cyber threats quarter 1 2020,
https://www.ptsecurity.com/ru-
ru/research/analytics/cybersecurity-threatscape-2020-
q1/, last accessed 2022/03/25.
21. Ship cybersecurity is in the hands of sailors,
https://moryakukrainy.livejournal.com/3843048.html,
last accessed 2022/03/25.
22. Vulnerabilities,
https://www.securitylab.ru/vulnerability/, last accessed
2022/03/25.