Abstract
Cryptography is a pivotal application of graph theory in ensuring secure communication systems. Modern cryptography is deeply rooted in mathematical theory and computer science practices. It is widely recognized that encryption and decryption processes are primarily outcomes of mathematical research. Given the increasing importance of safeguarding secret information or messages from potential intruders, it is imperative to develop effective technical tools for this purpose. These intruders are often well-versed in the latest technological advancements that could breach security. To address this, our study focuses on the efficacious combinatorial technique of graph networks using efficient domination and total magic labeling. The introduction of a graph network based on total magic labeling can significantly influence the network’s performance. This research introduces a novel combinatorial method for encrypting and decrypting confidential numbers by leveraging an efficient dominant notion and labeled graph.
MSC:
05C78; 05C69; 05C90
1. Introduction
Let represent an undirected connected graph with vertices and edges. The degree, neighborhood and closed neighborhood of a vertex in the graph are denoted by and , respectively. Ore [1] was the pioneer in introducing the terms “dominating set” and “dominating number” in his research. For a comprehensive understanding of domination theory-related terms, readers can refer to [2]. Numerous types of domination parameters have been introduced and explored over time. Teresa et al. [3] delved into various forms of dominations and their associated constraints. Cockayne et al. [4] delved into total dominance, a study further expanded in [5,6]. Harary and others explored double domination in [7], while studies on paired domination are documented in [8,9]. Chellali et al. [10] explored the relationship between total and paired domination in trees. Further insights on paired domination are available in [11,12,13]. Dunbar and Haynes [14] examined domination in inflated graphs, with their research continuing in [15,16,17,18]. Sampath Kumar initiated the study of equitable domination, which was later expanded upon by Swaminathan and Dharmalingam [19]. Meenakshi delved into the equitable domination of the complements of inflated graphs in [20,21].
Rosa [22] was the first to introduce graph labeling in 1967 and continued his studies on tree labeling [23]. Yeh embarked on a survey of labeling results for graphs with a distance of two. Grannell et al. [24] introduced modular gracious labeling of trees, a concept that bridges α-labeling and graceful labeling. Further studies on graceful labeling can be found in [25,26,27]. Prime labeling was initially explored by Tout, Dabboucy and Howalla [28] in 1982. Vertex equitable labeling was pioneered by Jeyanthi et al. [29], and various types of labeling were subsequently studied, as documented in [30,31,32,33,34,35,36]. It is important to note that our focus is solely on finite, simple and connected graphs.
The primary motivation behind this research is to devise novel algorithms impervious to intruders. It is essential for potential adversaries to understand the graphical network development statistics and the parameters used to decipher concealed sensitive data represented as numerical values [37,38,39]. This research introduces an encryption system comprising (i) Plain text, (ii) Encryption algorithm, (iii) Secret key, (iv) Cypher text and a decryption system consisting of (i) Cypher text, (ii) Secret Key, (iii) Decryption algorithm, (iv) Plain text. Initially, the general graph network (with modulo t) is constructed using the encryption algorithm, which leverages parameters (secret key), efficient domination and total magic labeling. The resulting graph (Cypher text) embeds hidden messages or information in the form of numerical values located in the incident links of the efficient dominating set members. To access this secret information, one must decrypt the provided algorithm. Examples are provided for the construction of graph networks with modulo 5 and modulo 7. Additionally, an adversary model has been established for this research, with discussions centered on various facets of the attacker [40,41].
Utilizing the aforementioned methodology, one can establish three distinct networks [42,43,44]: a graph network with a minimum number of edges, a graph network with a moderate number of edges and a graph network with a maximum number of edges [45,46,47]. Opting for the first network, characterized by the fewest edges, exposes the system to vulnerabilities, as adversaries can easily discern the concealed secret information, highlighting a significant system flaw [48,49,50]. In the subsequent discussion section, we will delve deeper into the proposed method, analyzing its strengths and weaknesses in comparison to existing approaches. This will provide readers with a comprehensive understanding of the method’s efficacy and potential areas of improvement. By juxtaposing our method with established techniques, we aim to elucidate its unique advantages and potential challenges, offering a holistic perspective on its applicability in real-world scenarios.
2. Preliminaries
Definition 1
([1]). Let be a finite, simple and connected graph. A set is a dominating set if, for every vertex in but not in , there exists at least one vertex in such that is adjacent to . The cardinality of the smallest dominating set in is termed the domination number of and is denoted by
Definition 2
([33]). A set is a total dominating set if every vertex in is adjacent to a vertex of . The cardinality of the smallest total dominating set is termed the total domination number and is denoted by
Definition 3
([34]). A dominating set is a paired dominating set if the induced subgraph has a perfect matching. The smallest cardinality of a paired dominating set in is termed the paired domination number of and is denoted by
Definition 4
([34]). A dominating set is termed an efficient dominating set ( ) if, for every vertex Equivalently, a dominating set is efficient if the distance between any two vertices in is at least three.
Definition 5
([24]). Let be a finite simple connected graph. Labeling of is a mapping from a set of vertices, edges or both to a set of labels. Typically, labels are positive (or non-negative), but real numbers can also be employed.
Definition 6
([24]). A function of to a set of labels is vertex labeling. A graph having such a function is called a vertex-labeled graph.
Definition 7
([29]). A function of to a set of labels is edge labeling. A graph having such a function is called an edge-labeled graph.
Definition 8
([29]). finite undirected graphs without loops and many edges with vertex set and edge set , where , . A graph labeling is any mapping that translates a certain set of graph elements to a specific set of positive integers. If the domain is a vertex (or edge) set, the labeling is referred to as vertex (or edge) labeling. When the domain contains both vertices and edges, the labeling is referred to as total labeling.
Definition 9
([29]). Graph labelings can assign weights to each edge and/or vertex. If every vertex weight (or edge weight, as appropriate) has the same value, the labeling is termed magic.
Definition 10
([27]). Let denote a graph with the vertex set and the edge set . A vertex magic total labeling of is a one-to-one mapping if there exists a constant c such that for every vertex x of . f . The magic constant for f is the constant c.
Definition 11
([27]). If there is a bijection such that there exists a constant “c” for any (l, m) in satisfying , then the graph with vertices and edges are said to be total edge magic.
3. Graph Network Construction
Let HGN = (VGN, EGN) be a graph that has
Let be the EeffDS members with degrees say , and , where .
Let ; be the neighboring nodes of respectively, …; Define and
Let for only one where
Let . Then
Define
Then . H′GN satisfies,
4. Encryption System of Graph Network
It consists of (i) Plain text, (ii) Encryption Algorithm, (iii) Secret Key, (iv) Cypher Text.
4.1. Plain Text
Plain text of the graph network is the secret information or secret messages in the form of numerical value using modulo t; t is a positive integer. Let the secret number be LSN ≡ asn (mod t), where asn = 0.
4.2. Encryption Algorithm: LSN ≡ asn (mod t), Where asn = 0
Input: , , where asn = 0.
Output: Encrypted Labeled Graph (ELG), H′GN
Start
Step 1: Divide LSN into t separations, say , where .
Step 2: Fix . Let ; be the neighboring nodes of respectively, …;
Step 3: Define and
Let for only one where Let
Now let
Then H′GN satisfies,
Step 4: …, .
Step 5: Split ; where k1, k2, …, kt
Step 6: Let defined as being a many-to-one function. Let where N* = {;
and such that for be a many-to-one function such that every edge is a constant. The constructed graph network preserves total magic labeling.
Step 7: Define
and
Define
End.
4.3. Secret Key
Find the Eeff DS of the graph network, which is .
4.4. Cypher Text
Using an encryption algorithm and secret key, we have constructed the sub-graph network GN1, GN2, GN3, …, GNt , which makes the labeled graph network. There are three possible instances considered here, according to the construction. (i) Labeled graph network with a minimum number of edges say H′GN min-(t), shown in Figure 1. (It is a labeled graph network tree.) (ii) Labeled graph network with a moderate number of edges say H′GN mod-(t), shown in Figure 2. (iii) Labeled graph network with more number of edges say H′GN more-(t), shown in Figure 3.
Figure 1.
H′GN min-(t).
Figure 2.
H′GN mod-(t).
Figure 3.
H′GN more-(t).
5. Decryption System of Graph Network
(i) Cypher text, (ii) Secret key, (iii) Decryption algorithm, (iv) Plain text.
Cypher text is the labeled graph network. Using a secret key, we decrypt it, and we get back the plain text.
Decryption Algorithm
Input: ELG, H′GN
Output: LGN, the plain text.
Begin
Step 1: Find the
such that
Step 2:
where
Step 3:
End.
6. Illustration: Secret Number LSN ≡ asn (mod t), Where asn = 0 and t = 5
6.1. Graph Network for LSN ≡ asn (mod t), Where asn = 0 and t = 5
Let HGN = (VGN, EGN) be a graph that has Let X = be Eeff DS members with degrees say and where
Let ; and be the neighboring nodes of respectively, .
Let for only one where
Let Then
Let , where . H’GN satisfies .
6.2. Encryption Algorithm
LSN ≡ asn (mod t), where asn = 0 and t = 5.
Input: , , where asn = 0
Output: ELG, H′GN
Start
Step 1: Divide LSN into t divisions say , where
Step 2: Fix where Let ; be the neighboring nodes of , respectively, and
Step 3: and
Let for only one where
Let
Then
Let
where
Then
where
H′GN satisfies
Step 4:
Step 5: Split
where k1, k2, k3, k4, k5
Step 6: Let defined as being a many-to-one function. Let be a many-to-one function where ; } and for such that every edge is a constant. The constructed graph network preserves total magic labeling.
Step 7: Define
End.
6.3. Decryption Algorithm
Input: ELG, H′GN
Output: LGN.
Begin
Step 1: Find the EeffDS, such that .
Step 2:
Step 3:
End.
6.4. Secret Number LSN = 12,935 (mod t), Where = 0 and t = 5
In this illustration, we split the LSN into seven parts and assign each to the EeffDS members and the connections arising on them, so that where and .
6.4.1. Encrypting LSN, 12,935
Step 1: Split LSN = 12,935 into five partitions say L0, L1, L2, L3 and L4, namely L0 = 2575 ≡ 0 (mod 5), L1 = 2581 ≡ 1 (mod 5), L2 = 2587 ≡ 2 (mod 5), L3 = 2593 ≡ 3 (mod 5) and L4 = 2599 ≡ 4 (mod 5).
Step 2: Construct the graph H′GN with , = 69 and its EeffDS, S’GN = .
Step 3: Arbitrarily choose
Let ; be the neighboring nodes of , respectively.
Step 4: Choose
Step 5: Split = 515 = 104 + 105 + 100 + 102 + 104 (where Split = 516 = 106 + 100 + 103 + 104 + 103 (where Split = 517 = 105 + 103 + 101 + 103 + 105 (where Split = 518 = 100 + 102 + 104 + 106 + 106 (where Split = 519 = 101 + 102 + 104 + 106 + 106 (where
Step 6: Let defined it as a many-to-one function. Let be a many-to-one function where } and such that for such that for every edge is a constant. The constructed graph network preserves magic labeling.
Step 7: Define
Define
Define
Define
Define
The encrypted labeled graph network for LSN ≡ asn (mod t), where asn = 0 and t = 5. Say H′GN-(5) is shown in Figure 4.
Figure 4.
H′GN-(5).
6.4.2. DECRYPTING LSN, 12,935
Step 1: Find the EeffDS, such that .
Step 2: = 5 (104 + 105 + 100 + 102 + 104) = 2575 = 1 + 5(106 + 100 + 103 + 104 + 103) = 2581 = 2 + 5(105 + 103 + 101 + 103 + 105) = 2587 = 3 + 5(100 + 102 + 104 + 106 + 106) = 2593 = 4 + 5(101 + 102 + 104 + 106 + 106) = 2599
Step 3: = 2575 + 2581 + 2587 + 2593 + 2599 = 12,935.
7. Illustration: Secret Number LSN ≡ asn (mod t), Where asn = 0 and t = 7
7.1. Graph Network Construction LSN ≡ asn (mod t), Where asn = 0 and t = 7
Let HGN = (VGN, EGN) be a simple graph with Let be the set of nodes, which are the members of Eeff DS with degrees say , , and where Let ; ; and be the neighboring nodes of respectively, .
Let for only one where Define
Then
Define , where Then
The edge set’s cardinality for the set of H′GN satisfies
We have presented the algorithm for creating a network and labeling its nodes and edges to encrypt the secret number LSN.
7.2. Encryption Algorithm
LSN ≡ asn (mod t), where asn = 0 and t =7.
Input: , , where asn = 0.
Output: ELG, H′GN
Begin
Step 1: Divide LSN into t divisions say where
Step 2: Fix where
Let ; … be the neighboring nodes of respectively, and .
Step 3: Define and
Let for only one where
Let
Then
Let
where Then , where
H’GN satisfies
Step 4:
Step 5: Split
where k1, k2, k3, k4, k5, k6, k7
Step 6: Let defined as being a many-to-one function. Let be a many-to-one function where ; } and such that for such that for every edge is a constant. The constructed graph network preserves total magic labeling.
Step 7: Define
End.
7.3. Decryption Algorithm
Input: ELG, H′GN
Output: LGN
Begin
Step 1: Find the Eeff DS such that .
Step 2:
Step 3:
End.
7.4. ILLUSTRATION: Secret Number LSN = 35497 ≡ asn (mod t), Where asn = 0 and t = 7
7.4.1. Graph Network Construction for LSN ≡ asn (mod t), Where asn = 0 and t = 7
In this example, we divide LSN into seven parts and allocate each to the Eeff DS members and the connections arising on them, so that where and .
7.4.2. ENCRYPTING LSN, 35,497
To encrypt the confidential number 35,497 which is a multiple of 7, let LSN = 35,497 ≡ asn (mod t), where asn = 0 and t = 7.
Step 1: Split LSN = 35,497 into seven partitions say L0, L1, L2, L3, L4, L5 and L6 namely L0 = 5047 ≡ 0 (mod 7), L1 = 5055 ≡ 1 (mod 7), L2 = 5063 ≡ 2 (mod 7), L3 = 5071 ≡ 3 (mod 7) and L4 = 5079 ≡ 4 (mod 7), L5 = 5087 ≡ 5(mod 7), L6 = 5095 ≡ 6 (mod 7).
Step 2: Arbitrarily choose deg(m1) = 4, deg(m2) = 5, deg(m3) = 6, deg(m4) = 6, deg(m5) = 5, deg(m4) = 7 and deg(m5) = 7. Let ; ; … be the neighboring nodes of , respectively.
Step 3: Construct the graph H′GN with and = 81 and its Eeff DS, S’GN = {m1, m2, m3, m4, m5, m6, m7}.
Step 4: ,
Step 5: Split = 721 = 125 + 188 + 164 + 244 (where Split = 722 = 195 + 213 + 103 + 102 + 109 (where
Split = 723 = 176 + 104 + 135 + 102 + 100 + 106 (where
Split = 724 = 105 + 104 + 103 + 102 + 160 + 150 (where
Split = 725 = 105 + 144 + 153 + 148 + 175 (where
= 726 = 106 + 105 + 104 + 103 + 102 + 100 + 106 (where
= 727 = 103 + 102 + 101 + 106 + 106 + 105 + 104 (where
Step 6: Let defined as being a many-to-one function. Let be a many-to-one function where ; ; and such that for such that , for every edge is a constant. The constructed graph network preserves total magic labeling.
Step 7: Define
Define
Define
Define
Define
Define
Define
Figure 5.
H′GN-(7).
7.4.3. DECRYPTING LSN, 12,935
Step 1: Find the Eeff DS, such that
Step 2:
Step 3: = 5047 + 5055 + 5063 + 5071 + 5079 + 5087 + 5095 = 35,497.
8. Adversary Model
In the realms of cybersecurity, cryptography and security analysis, an adversary model—often termed a threat model—is a conceptual framework designed to characterize and delineate potential attackers or threats that a system, protocol or application might face. This model assists in identifying and understanding the myriad strategies an adversary might employ to compromise a system’s security [51,52,53].
Capabilities: This aspect pertains to the abilities of an attacker in terms of resources, knowledge and access. The spectrum of capabilities can span from rudimentary attacks by novices, often called “script kiddies”, to intricate assaults orchestrated by well-funded and adept hackers. Our cryptosystem, grounded in a network construction based on modulo t, total labeling and efficient domination, hinges on knowledge-based resources. If an intruder is privy to this information, they can evolve into proficient attackers of this system.
Goal: This dimension focuses on the objectives of the attacker. Potential aims might encompass data theft, service disruption, user impersonation or other malicious endeavors. The primary objective of adversaries targeting this cryptosystem is to decipher the concealed information.
Knowledge: This refers to the attacker’s understanding of the system, protocol or technology they aim to exploit. Such knowledge encompasses awareness of potential weaknesses, entry points and vulnerabilities. In the context of our cryptosystem, knowledge is anchored in two pivotal parameters: efficient domination and total labeling. If the constructed network exhibits minimal connectivity, attackers can more readily pinpoint the nodes associated with efficient domination.
Constraints: These represent the limitations or challenges an attacker might encounter. A quintessential constraint might be the attacker’s finite processing power or a narrow window of opportunity to strike. Within our cryptosystem, certain constraints are inherent, such as the necessity to select an appropriate secret number aligned with modulo t, partitioning this secret number and utilizing the partitions to assign weights to the network’s edges.
Resources: This facet delves into the arsenal at the attacker’s disposal, which could encompass financial resources, computational prowess, network infrastructure and access to specialized tools or software. In the context of our cryptosystem, the network’s construction hinges on the analysis of a secret number. An astute attacker might deduce that the concealed information is encapsulated as a numerical value.
9. Conclusions
We have presented a combinatorial approach to ascertain the secret number as manifested in the efficiently constructed graph network. The principles of total magic labeling and efficient domination are pivotal in crafting such an efficient graph network. The key to this system lies in identifying the most efficient and dominant set, termed . By assigning repeated values of and considering the maximum number of edges present in the constructed , the network’s complexity is heightened, making it more challenging to decode.
Author Contributions
Conceptualization, A.M. and A.K.; Data curation, A.M. and M.E. Formal analysis, A.M. and R.C.; Investigation, A.M., R.C. and A.K.; Methodology, A.M. and M.E. Supervision, A.M., M.E. and R.C.; Visualization, A.M. and A.K. Writing—original draft, R.C., A.M. and A.K. Writing—review and editing, A.M., A.K. and R.C. All authors have read and agreed to the published version of the manuscript.
Funding
This work was supported by the project SP2023/088 Application of Machine and Process Control Advanced Methods supported by the Ministry of Education, Youth and Sports, Czech Republic.
Data Availability Statement
The data availability of this research work is based on the published research article of the relevant topics.
Conflicts of Interest
The authors declare no conflict of interest.
References
- Ore, O. Theory of Graphs; American Mathematical Society Colloquium Publications: Providence, RI, USA, 1962. [Google Scholar]
- Bondy Murthy, J.A. Graph Theory with Applications; Elsevier Science Publishing Co., Inc.: New York, NY, USA, 1976. [Google Scholar]
- Haynes, T.W.; Hedetniemi, S.T.; Slater, P.J. Fundamentals of Domination in Graphs; Marcel Decker: New York, NY, USA, 1998. [Google Scholar]
- Cockayne, E.J.; Dawes, R.M.; Hedetniemi, S.T. Total domination in graphs. Networks 1980, 10, 211–219. [Google Scholar] [CrossRef]
- Bujtás, C. Domination number of graphs with minimum degree five. Discuss. Math. Graph Theory 2021, 41, 763–777. [Google Scholar] [CrossRef]
- Favaron, O.; Kabanov, V.; Puech, J. The ratio of three domination parameters in some classes of claw-free graphs. J. Comb. Math. Comb. Comput. 1999, 31, 151–160. [Google Scholar]
- Harary, F.; Haynes, T.W. Double domination in Graphs. Ars Comb. 2000, 55, 201–213. [Google Scholar]
- Gorzkowska, A.; Henning, M.A.; Kleszcz, E. Monika Pilśniak, Paired Domination in Trees. Graphs Comb. 2022, 38, 129. [Google Scholar] [CrossRef]
- Henning, M.A.; Pilśniak, M.; Tumidajewicz, E. Bounds on the paired domination number of graphs with minimum degree at least three. Appl. Math. Comput. 2022, 417, 12782. [Google Scholar] [CrossRef]
- Chellali, M.; Haynes, T.W. Total and paired domination numbers of a tree. AKCE Int. J. Graphs Comb. 2004, 1, 69–75. [Google Scholar]
- Edwards, M. Critically Concepts for Paired Domination in Graphs. Master’s Thesis, University of Victoria, Victoria, BC, Canada, 2006. [Google Scholar]
- Brigham, R.C.; Chinn, P.Z.; Dutton, R.D. Vertex domination-critical graphs. Networks 1988, 18, 173–179. [Google Scholar] [CrossRef]
- Hou, X.; Edwards, M. Paired Domination Vertex Critical Graphs. Graphs Comb. 2008, 24, 453–459. [Google Scholar] [CrossRef]
- Dunbar, J.E.; Haynes, T.W. Domination in inflated graphs. Congr. Numer. 1996, 118, 143–154. [Google Scholar]
- Favaron, O. Irredundance in inflated graphs. J. Graph Theory 1998, 28, 97–104. [Google Scholar] [CrossRef]
- Puech, J. The lower irredundance and domination parameters are equal for inflated trees. J. Comb. Math. Comb. Comput. 2000, 33, 117–127. [Google Scholar]
- Puech, J. Lower Domination Parameters in Inflated Trees; Research Report 97-57; Mathematical Department, UniversitNe Paris-Sud: Paris, France, 1997. [Google Scholar]
- Henning, M.A.; Kazemi, A.P. Total domination in inflated graphs. Discret. Appl. Math. 2012, 160, 164–169. [Google Scholar] [CrossRef][Green Version]
- Swaminathan, V.; Dharmalingam, K.M. Degree equitable domination on graphs. Kragujev. J. Math. 2011, 35, 191–197. [Google Scholar]
- Meenakshi, A. Equitable domination of complement of inflated graph. AIP Conf. Proc. 2019, 2112, 020079–1–6. [Google Scholar]
- Kumaran, N.; Meenakshi, A.; Mahdal, M.; Prakash, J.U.; Guras, R. Application of Fuzzy Network Using Efficient Domination. Mathematics 2023, 11, 2258. [Google Scholar] [CrossRef]
- Rosa, A. On Certain Valuations of the Vertices of a Graph, Theory of Graphs; Internat. Symposium: Rome, Italy, 1967. [Google Scholar]
- Hung, C.; Kotzig, A.; Rosa, A. Furter results on tree labellings. Util. Math. 1982, 21, 31–48. [Google Scholar]
- Grannell, M.J.; Griggs, T.S.; Holroyd, F.C. Modular gracious labellings of trees. Discret. Math. 2001, 231, 199–219. [Google Scholar] [CrossRef]
- Slater, P.J. On k-graceful graphs. In Proceedings of the 13th Southeastern Conference on Combinatorics, Graph Theory, and Computing, Boca Raton, FL, USA, 15–18 February 1982; Volume 53, p. 57. [Google Scholar]
- Lo, S.P. On edge-graceful labelings of graphs. Congr. Numer. 1985, 50, 231–241. [Google Scholar]
- MacDougall, J.A.; Miller, M.; Wallis, W.D. Vertex-magic total labelings of graphs. Util. Math. 2002, 61, 3–21. [Google Scholar]
- Tout, A.; Dabboucy, A.N.; Howalla, K. Prime labeling of graphs. Nat. Acad. Sci. Lett. 1982, 11, 365–368. [Google Scholar]
- Jeyanthi, P.; Maheswari, A. Some results on vertex equitable labeling. Open J. Discret. Math. 2012, 2012, 18867. [Google Scholar] [CrossRef][Green Version]
- Zhang, X.; Ye, C.; Zhang, S.; Yao, B. Graph Colorings and Labelings Having Multiple Restrictive Conditions in Topological Coding. Mathematics 2022, 10, 1592. [Google Scholar] [CrossRef]
- Zhang, X.; Ibrahim, M.; Bokhary, S.A.U.H.; Siddiqui, M.K. Edge irregular reflexive labeling for the disjoint union of gear graphs and prism graphs. Mathematics 2018, 6, 142. [Google Scholar] [CrossRef]
- Hao, J.; Gong, Y.; Sun, J.; Tan, L. Use the K-Neighborhood Subgraphs to Compute Canonical Labelings of Graphs. Mathematics 2019, 7, 690. [Google Scholar] [CrossRef]
- Allan, R.B.; Laskar, R.; Hedetniemi, S. A note on total domination. Discret. Math. 1984, 49, 7–13. [Google Scholar] [CrossRef]
- Liu, X.; Shi, T.; Zhou, G.; Liu, M.; Yin, Z.; Yin, L.; Zheng, W. Emotion classification for short texts: An improved multi-label method. Humanit. Soc. Sci. Commun. 2023, 10, 306. [Google Scholar] [CrossRef]
- Liu, X.; Zhou, G.; Kong, M.; Yin, Z.; Li, X.; Yin, L.; Zheng, W. Developing Multi-Labelled Corpus of Twitter Short Texts: A Semi-Automatic Method. Systems 2023, 11, 390. [Google Scholar] [CrossRef]
- Meenakshi, A.; Mythreyi, O.; Bramila, M.; Kannan, A.; Senbagamalar, J. Application of neutrosophic optimal network using operations. J. Intell. Fuzzy Syst. 2023, 45, 421–433. [Google Scholar] [CrossRef]
- Cheng, B.; Zhu, D.; Zhao, S.; Chen, J. Situation-Aware IoT Service Coordination Using the Event-Driven SOA Paradigm. IEEE Trans. Netw. Serv. Manag. 2016, 13, 349–361. [Google Scholar] [CrossRef]
- Li, B.; Zhou, X.; Ning, Z.; Guan, X.; Yiu, K.C. Dynamic event-triggered security control for networked control systems with cyber-attacks: A model predictive control approach. Inf. Sci. 2022, 612, 384–398. [Google Scholar] [CrossRef]
- Zhuang, Y.; Jiang, N.; Xu, Y.; Xiangjie, K.; Kong, X. Progressive Distributed and Parallel Similarity Retrieval of Large CT Image Sequences in Mobile Telemedicine Networks. Wirel. Commun. Mob. Comput. 2022, 2022, 1–13. [Google Scholar] [CrossRef]
- Idrees, M.S.; Roudier, Y.; Apvrille, L. Model the system from adversary viewpoint: Threats identification and modeling. Adv. Intrusion Prev. Workshop 2014, 165, 45–58. [Google Scholar]
- Zhang, J.; Peng, S.; Gao, Y.; Zhang, Z.; Hong, Q. APMSA: Adversarial Perturbation Against Model Stealing Attacks. IEEE Trans. Inf. Forensics Secur. 2023, 18, 1667–1679. [Google Scholar] [CrossRef]
- Jin, H.; Wang, Z.; Wu, L. Global dynamics of a three-species spatial food chain model. J. Differ. Equ. 2022, 333, 144–183. [Google Scholar] [CrossRef]
- Liu, P.; Shi, J.; Wang, Z.-A. Pattern formation of the attraction-repulsion Keller-Segel system. Discret. Contin. Dyn. Syst.-Ser. B 2013, 18, 2597–2625. [Google Scholar] [CrossRef]
- Jin, H.Y.; Wang, Z.A. Global stabilization of the full attraction-repulsion Keller-Segel system. Discret. Contin. Dyn. Syst.-Ser. A 2020, 40, 3509–3527. [Google Scholar] [CrossRef]
- Wang, F.; Wang, H.; Zhou, X.; Fu, R. A Driving Fatigue Feature Detection Method Based on Multifractal Theory. IEEE Sens. J. 2022, 22, 19046–19059. [Google Scholar] [CrossRef]
- Yang, S.; Li, Q.; Li, W.; Li, X.; Liu, A. Dual-Level Representation Enhancement on Characteristic and Context for Image-Text Retrieval. IEEE Trans. Circuits Syst. Video Technol. 2022, 32, 8037–8050. [Google Scholar] [CrossRef]
- Liu, A.-A.; Zhai, Y.; Xu, N.; Nie, W.; Li, W.; Zhang, Y. Region-Aware Image Captioning via Interaction Learning. IEEE Trans. Circuits Syst. Video Technol. 2022, 32, 3685–3696. [Google Scholar] [CrossRef]
- Qiao, F.; Li, Z.; Kong, Y. A Privacy-Aware and Incremental Defense Method Against GAN-Based Poisoning Attack. IEEE Trans. Comput. Soc. Syst. 2023. [Google Scholar] [CrossRef]
- Jiang, H.; Wang, M.; Zhao, P.; Xiao, Z.; Dustdar, S. A Utility-Aware General Framework with Quantifiable Privacy Preservation for Destination Prediction in LBSs. IEEE/ACM Trans. Netw. 2021, 29, 2228–2241. [Google Scholar] [CrossRef]
- Ma, J.; Hu, J. Safe consensus control of cooperative-competitive multi-agent systems via differential privacy. Kybernetika 2022, 58, 426–439. [Google Scholar] [CrossRef]
- Tan, J.; Jin, H.; Hu, H.; Hu, R.; Zhang, H.; Zhang, H. WF-MTD: Evolutionary Decision Method for Moving Target Defense Based on Wright-Fisher Process. IEEE Trans. Dependable Secur. Comput. 2022. [Google Scholar] [CrossRef]
- Lu, S.; Liu, M.; Yin, L.; Yin, Z.; Liu, X.; Zheng, W. The multi-modal fusion in visual question answering: A review of attention mechanisms. PeerJ Comput. Sci. 2023, 9, e1400. [Google Scholar] [CrossRef]
- Lu, S.; Ding, Y.; Liu, M.; Yin, Z.; Yin, L.; Zheng, W. Multiscale Feature Extraction and Fusion of Image and Text in VQA. Int. J. Comput. Intell. Syst. 2023, 16, 54. [Google Scholar] [CrossRef]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).