INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL
Online ISSN 1841-9844, ISSN-L 1841-9836, Volume: 17, Issue: 6, Month: December, Year: 2022
Article Number: 4363, https://doi.org/10.15837/ijccc.2022.6.4363

CCC Publications 

Bi-Level Minimal Resource Protected Key Generation Framework For
Fog Computing Applications

Arul Sindhia.P, Bharathi.R

Arul Sindhia.P
University College of Engineering Nagercoil, 629004, India
Corresponding author: arulsindhiaphd@gmail.com

Bharathi.R
University College of Engineering Nagercoil, 629004, India.
BharathiBharathi1993@hotmail.com

Abstract

Fog computing is a viewpoint that expands on the Cloud stage concept by placing processing
assets at the organization’s edges. It might be described as a cloud-like platform with compara-
ble data, computation, storage, and applications. They are unique in that they are decentralized
in nature. Data protection and route analysis of time-sensitive data are made easier using fog
computing. This minimizes the volume and distance of data sent to the cloud, lowering the risk
of security and privacy breaches in IoT applications. When it comes to security and privacy, fog
computing confronts several issues. The constraints of fog computing resources are the root of these
difficulties. The fog system, in fact, may raise new security and privacy concerns. To address these
challenges, cryptography is used in conjunction with key management techniques to provide safe
data transfer. A Minimal Resource Viterbi based Bi-level Secured Key Generation (MRV-BSKG)
technique for a secured fog-based system is proposed to compromise the security level and com-
putational complexity. The BSKG technique, which combines Lagrange’s Key Generation (LKG)
and the Location-Based Key (LBK) generation approaches, can safeguard secrecy and integrity.
In comparison to the previous techniques, the new MRV, BSKG, delivers security with improved
outcomes.

Keywords:Minimal Resource Viterbi based Bi-level Secured Key Generation (MRV-BSKG),
Fog Computing, Lagrange’s Key, Location-Based Key, Shortest path.

1 Introduction
Fog computing is in great demand in most real-time applications [1], including industrial automa-

tion systems, smart grid systems, smart cities, smart buildings, health care systems, surveillance and
monitoring, traffic control systems, smart agriculture, and smart factories, among others. The Fog
System platform is divided into three layers: the end user layer, the fog layer, and the cloud layer
[2]. Fog System serves as a link between IoT devices and cloud computing. The fog computer gadget
may send many types of new information. Cloud computing and fog computing have similar working



https://doi.org/10.15837/ijccc.2022.6.4363 2

principles. Nonetheless, with cloud computing, the cloud can handle massive volumes of data utilizing
virtual cloud resources which can be recovered from the cloud when necessary. The probability of
network traffic due to a huge quantity of data transfer presents a key barrier in cloud computing
[3]. The data transfer between the smart devices and the cloud requires adequate time and appro-
priate bandwidth. However, maintaining these standards during data transfer in cloud computing is
difficult. There are additional issues with latency, location awareness, and mobility [4]. The Inter-
net of Things (IoT) is becoming an inextricable aspect of the Internet’s expanding privacy inside its
bounds. However, IoT-based gadgets and tools face security and privacy risks, particularly because
nodes are primarily meant to collect data about users’ behaviours, keys, and surroundings, resulting
in profitable, targeted invasions. CISCO has launched Fog Computing, an improved version of cloud
computing, to address these concerns. Fog nodes are employed at the network’s edge and act as a link
between smart devices and the cloud. Smart population of devices are linked to fog edges rather than
having direct touch with the cloud. The solution is ideal for any latency sensing applications without
smart devices that can quickly obtain data thanks to fog computing. As a result, an approximation
approach that provides scaled graphics processing with substantial acceleration is required. A vector-
based machine learning algorithms is used [21] to approximate the shortest path distance in a large
image. The main focus of the fog system is security, which can be achieved using fog forecasts and
several key management schemes [5].It should be considered that the design of security methods has
no impact on the system’s performance. Fog-based system design with less security leads to fraudulent
activities such as hacking of information about the users. Encryption technology will also be used to
improve the security of fog-based systems.But the ineffective key management technique affects the
communication between the user and the cloud [6]. So there is a demand for a highly secured Fog
based system to make it suitable for time-sensitive applications. In this paper, fog based system with
Bi-level Secured Key Generation (BSKG) method is proposed that combines Lagrange’s Key Genera-
tion (LKG) and Location-Based Key (LBK) generation. During the data transmission from source to
fog nodes, LKG is done using a source encoder, and in every fog node, LBK is done using a channel
encoder. The generated key is a combination of alphabetical letters and numerical characters leading
tohigh-level security. On the receiver side, the generated key is decrypted using the LBK method in
the channel decoder and again decrypted using the LKG method in the source decoder and then stored
in the cloud. Here, the information is secured in multiple levels by encoding at source and each node
based on its position. This approach provides a common security mode and provide secured data as
compared to other algorithms like attribute based encryption or machine learning model. The rest of
the paper is as follows. Existing works relating to techniques for fog computing and key management
are described in Section 2.The fog-based smart grid system model is presented in Section 3.Section 4
describes the structure of the proposed fog-based system for the main generation of two-level protec-
tion. Section 5 describes the functions of the proposed smart grid system, and Section 6 concludes
the work.

2 Related work
Cloud computing and fog computing services provide the same data processing, from consumers to

these collectors, which are then stored in the cloud [7, 8].Cloud computing is a three-level structure,
while fog computing is a tri-level structure.This tri-level structure supports real time applications
due to its better services in terms of Quality of Service (QoS), latency, geographical distribution and
network traffic [9].To reduce the latency in fog computing, shortest path algorithms such as Dijikshtra,
Thorup’s algorithm, Pulse Coupled Neural Network (PCNN), Time Delay Neural Network (TDNN),
etc. are discussed by Huang et al. which determines the minimum distance between the source and
destination with the neural network. The main thing not to pay attention to in fog estimates is
security and privacy issues. Only after the fog nodes collect sensitive information about the user will
we send it to the cloud server.So, there may be a possibility for security threats [10].Key management
techniques and cryptography algorithms can improve securities to fog nodes and cloud servers. Mate
Horwath [11] has proposed Attribute-Based Encryption (ABE) that uses Linear Secret Sharing Scheme
(LSSS),which supports multiple users. It is an identity-based user revocation that reduces the cloud’s



https://doi.org/10.15837/ijccc.2022.6.4363 3

computational overhead by increasing the computational overhead in both encryption and decryption.
The modified structure of Ciphertext Policy ABE (CP-ABE) is presented in [12] that effectively
exchanges the hierarchy files in the cloud to reduce the time required for decryption. Simultaneously,
the user can get all files by using a single key, which reduces the overall system’s cost. Peng Zhang et al.
proposed a CP-ABE scheme to reduce the cloud’s burden by transferring some computation processes
to fog nodes [13]. In this structure, the number of attributes is independent of the encryption and
decryption process. Key Management Scheme for Communication Layer (KMS-CL)method exploits
an access control approach to improve the security during file transmission in fog nodes. It increases
the computational complexity due to the outsourced encryption process. To provide better security
and privacy services, privacy-preserving data aggregation methods are presented in [14-16]. In [11-
13], ABE based mechanisms only were used for securing the data. The data can be hacked if the
user information is leaked or hacked by third party. But, in [14 -16], this problem is overcome by
using encryption mechanism; but its information is processed through third party and it also risk
of information leakage. Lightweight Privacy Data Preserving Aggregation (LDPA) was proposed
by Rongxing Lu et al. [17], primarily for IoT applications. To provide great security and privacy,
it employs the Chinese Remainder Theorem (CRT) with Parlier encryption. The notion of shared
keys is a flaw in this strategy since it cannot guarantee collision resistance. For traditional systems,
different processing and machine learning modules are used in the physiological data processing cloud
[23].The machine learning model [22 -23] based processing is best for only particular model and
achieves best security in that model alone. Tian Wang et al. [18] have presented a trajectory privacy
preservation method that uses Dummy Rotation (DR) algorithm to provide location-based services.
By using this approach, privacy preservation is achieved, but it doesn’t focus the integrity. Zhi Li et al.
presented a novel Hyper-graph-based Key Management (HKM) scheme that focused on confidentiality
and integrity [19]. It has a key generation center for generating secret keys, fog server for processing
the data, cloud for data storage and users. The scheme secures the storage data in both forward
and backward directions. Bashar Alohali et al. [20]have presented the key management scheme in
the Smart Grid (KM-CL-SG) communication layer. Since the security and privacy issues threaten
modern technology, various researches focus on achieving high security.

3 Implementation of the Proposed System
The major problems presented in the cloud-based system are data overloading and latency that

lead to user data hacking. This can be overcome by fog-based organizations that play a key role in the
construction of smart cities and smart transportation. The system model of the proposed fog-based
system is shown in Figure 1. The application layer, network layer, and perception layer are the three

Figure 1: The system model of Fog based System

levels that make up FOG system. The perception layer at the bottom of the design is where the IoT



https://doi.org/10.15837/ijccc.2022.6.4363 4

devices are positioned. Because every IoT node senses and collects the users’ information for further
processing, it’s also known as an end-user layer. The network layer, which sits in the middle of the
architecture, serves as a link between the perception and application levels. For processing and sending
data to the upper layer, it incorporates fog nodes and network components such as routers, gateways,
switches, and so on. The data are saved in the cloud at the top layer, which is an application layer.

3.1 Proposed Fog based System

Due to the growth in population and technology, the Fog-based system is necessary to communicate
between the cloud server and customers. The block diagram of the proposed fog-based system is shown
in Figure 2.

Figure 2: Proposed System Diagram

While Fog Node offers services used in different locations, cloud computing tipping offers its own
cache and subsequent requests locally. Act as the central controller of global service delivery and
distributed fog nodes. In addition, the cloud image receives the necessary information from the fog
population sensor on the end user device, which can now communicate with the IoT sensors central
information database, and then use the fog terminal to communicate. It not only provides information
about the devices, but it also provides the user’s private information. So there is a possibility for
hacking or spoofing the user information. The proposed hybrid Bi-level can mitigate its secured
key generation scheme, which combines Lagrange’s Key Generation (LKG) and Location-Based Key
(LBK) generation.

The encryption procedure for safeguarding information about and from users is conducted out
during data transfer from end-users to fog nodes using Lagrange’s key generation technique. The fog
node stores the encrypted data created. In order to offer two-level security, LBK is created in every
fog node and combined with the encrypted data. Using MRV cloud nodes and fog nodes, reduce short
path secretion between users. The data is decrypted again on the cloud server using the encryption
logger key, and the data is eventually saved in the cloud, as shown in Figure 3.
Algorithm Step 1: Create n number of nodes n1,n2,n3, .....nn
Step 2: Assign the position of nodes and set the distance between nodes
for (i=0; i<n ;i++) (tmp =distance between node i and i+1............ (1) )
Step 3: Selection of source node (Si) and destination node (Di)
Step 4: Data transmission from Si to fog node (fi)
Step 5: Attacker determination by sending hello message to neighbor nodes
Step 6: Computation of shortest path using MRV by ignoring the attacker nodes
Step 7: BSKG key encryption LKG & LBK, LKG can be generated using equation 5, and LBK can
be generated based on the location of fog nodes
Step 8: Data transmission from fn to Dn



https://doi.org/10.15837/ijccc.2022.6.4363 5

Figure 3: Architecture Diagram of Data transmission with security

Step 9: BSKG key decryption LBK & LKG
Step 10: Detection of the mobile station for specific users
Step 11: Data reception in the Dn or cloud

3.2 Bi-level Secured Key Generation (BSKG)

The major concern in most modern applications is high-level security. By focusing on this, BSKG
is introduced, which is a combination of LKG and LBK key generation methods. It can generate
keys in terms of alphanumeric, which is a combination of alphabets and numbers. This effective key
generation method is applied when transmitted from the source to the fog node.

3.2.1 Lagrange’s Key Generation (LKG))

The data transmission from source to fog node generates a secret key based on Lagrange’s equation
to secure the customer data. If the systemyis represented as a function of x and it is defined in terms
of the parameter α, then

y = x + αψ + (y) (1)

Where α denotes Lagrange’s parameter, and ψ + (y) denotes the holomorphic function defined as a
complex differentiable function.
The following Lagrange’s expansion can compute the Lagrange’s key of the system y
Lagrange’s key generation approach uses the interpolation methods for performing both encryption
and decryption processes.Itapplies the interpolation to generate the polynomial for the data sensed
from the smart devices. The polynomial functions are generated between the smart devices and fog
nodes, and the data are represented as (x0, y0), (x1, y1),. . . . . . (xn,yn). It resulted in the nth degree
polynomial by utilizing n+1 data. The following equation can represent Lagrange’s polynomial with
the interpolation method.

f(y) = f(x) + α/1! ψ(x)f‘(x) + α2/2! ∂/∂x[ψ(x)]2f‘(x) + −−−−−− +αn+1[ψ(x)]n+1f‘(x) (2)



https://doi.org/10.15837/ijccc.2022.6.4363 6

Where n denotes the degree of polynomial and can be represented as follows

pn(x) =
n∑

j=0
Lj (x)f(xj ) (3)

The simplified form of (5) can be expressed as follows.

Lj (x) = x−x0/xj −x0.x−x1/xj −x1......x−xj−1/xj −xj−1......x−xn/xj −xn (4)

If the value of the x is known, then only the data is hacked. Still, it is difficult to predict the value
because it was chosen randomly ensures the security and confidentiality of the customer data.

3.2.2 Location-based Key Generation (LBK)

To attain secure and safe communication in fog computing, it should be ensured that secured
data is shared between fog nodes and the cloud. LBK key generation is applied in every fog node
after performing the encryption using Lagrange’s expansion. This can provide bi-level security for the
customer data protection. The key generated from LBK is an alphanumeric code that differs in every
fog node to achieve high-level security.It uses a novel key updating process depending on the grid’s
location information to avoid data corruption and hacking.

Lj (x) =
n∐

i=0
,J = 0, 1......n : i 6≡ j (5)

The working principle of the LBK key generation method is shown in Figure 4. This figure shows
securing of data against the attackers while the adversary tries to hack the data. It provides high-level
security by generating the key based on the location. It differs from the cryptographic key by its
alphanumeric key generation. In cryptography keys, the keys are generated as 32-bit binary keys. In
contrast, in LBK, the key is a combination of alphabets and numerals, which leads to high security
over other key management techniques.It is termed geo encryption because it is location-dependent,



https://doi.org/10.15837/ijccc.2022.6.4363 7

and it enables the routes for transmission from the source node to the destination node. If any attacker
is found in any of the nodes, the data is transferred through some other nodes. Since the data are
location authenticated, it can be transferred with high security.
To calculate each neighbor node location authenticated based average distance analysis

Figure 4: Location-Based Key Generation

Every fog node in the network generates a distinct key, and it is in the form of alphabets. After
performing LBK, the resultant key is alphanumeric. The key obtained from the LKG approach is
added in every fog node, increasing the security of the data transmission. The LPK system attempts
to locate the source / target by directly measuring the radio signal traveling between the transmitter
and the receiver. The relationship between the limit difference between the recipients and the LKG
is provided

Lj =
√

(x1 −y1)2 + (x2 −y2)2 (6)

Rm,n = s∗dm,n = Rm −Rn (7)

Where
s- Signal propagation speed
m,n – between neighbor node
R- Allocation of signal
These LBK, LKG systems use time, phase, or frequency measurement values to estimate the direction
or range of the first signal propagation path, and then use solutions that provide calculations based
on the distance of the data node measured by the generated key.

3.3 Minimal Resource Viterbi

In the proposed Fog-based Smart Grid system, information exchange with high speed is desirable.
This can be realized by finding the short path fog tip and fog tip cloud computing among users.MRV
is used to compute the shortest path effectively by exploiting dynamic programming approaches [21].
Instead of differential equations, minimizers are used in every neuron of this network. It uses adders
and minimizers for shortest path computation leading to low computational complexity. This reduces
the number of computations in each neuron and reduces the number of repetitions.This network
initiates the operation by activating the source neuron,which results in an auto wave, and then it
travels through all nodes of the network; finally, the best path is stored in the output.
Algorithm Steps Consider a graph G(N, V) with N nodes and V edges. Each node is assigned a
time window WT [li,ui] with upper(ui) and lower (li) threshold values. SPi is the shortest path from



https://doi.org/10.15837/ijccc.2022.6.4363 8

s to i, and TTi is the traversal time of the path to reach node i.
Step 1: Initialize the network, set the upper threshold and lower threshold in the threshold window

T LS = 0 : T
S
U = ε : T

L
i = li; Ti

U = ui + ε (8)

Step 2: Source neuron is activated by setting the input of the neuron within the threshold so that
the auto wave is generated and propagates to neighborhood neurons
Step3: Communication Range: To use the communication range of the behind the formula:

E = (u,v)�V 2|u 6= v (9)

Step4:Since the length of the shortest path from node ni
Step5: An available distance of node ni and required the minimum distance of node mi
Step6: Critical sensing point of the ni . If source node denote as and the nearest neighbor node
denote nnni<mi
Step 7: Competition of all minimum distance paths to the node happens, and this takes place
simultaneously in all neurons

Ci(t) = j,j�BS(i)andTTj (t) + dij [TTj (t)]�[TiL + TiU] (10)

Step 8: Compute traversal time at the output of the neuron,

TTiU (t + 1) = minTTi(t + 1),TTiU (t) (11)

Compute the shortest path from Nsto Ni Where l = winning node
Step 9: Update the upper threshold of the node
Step 10: Repeat the autosave travel and competition (step3 to step6) until the network converges.

When the MRV is used for the first time, the node will be removed, but the retained path infor-
mation will pass through the node just like all the previously mentioned paths; Therefore, when the
next search is made to proceed to this node, the stored path information will be used, and the map
can be easily used to find the shortest path at the fastest time to avoid recounting.

4 Results and Discussions
This chapter discusses the comparison of Minimal Resource Viterbi based Bi-level Secured Key

Generation (MRV-BSKG) and previous security;Bi-level Secured Key Generation provides authenti-
cate for the data when source node send the data and it is transmitted through the neighbouring node
in communication. The Minimal Resource Viterbichooses a different path if the neighbouring terminal
changes its position. So these two proposed methods are compared in this section.The previous meth-
ods Attribute-Based Encryption (ABE), Ciphertext Policy-ABE (CP-ABE), Lightweight privacy data
Preserving Aggregation (LDPA), Hyper-graph based Key Management (HKM) and Key Management
Scheme for Communication Layer (KMS-CL). The proposed Fog-based system’s performances are an-
alyzed in terms of packet drop, packet delivery ratio, throughput, and security.The parameters used
for simulation are tabulated in Table 1.

4.1 Packet Delivery Ratio (PDR)

It is defined as the ratio of the total number of packets received by the receiver atthe destination
to the total number of packets sent by the sender from the source.

PDR = PD/Ps (12)

Using the above equation 11, the PDR for the proposed scheme is compared with the existing ap-
proaches. Due to the highly secure key management mechanism and MRV approach, the proposed
fog-based system achieves a high PDR. Every path’s journey is saved in the hidden layers of MRV,



https://doi.org/10.15837/ijccc.2022.6.4363 9

Table 1: Simulation Parameters
Simulation Parameters Value

Number of nodes 50
Routing Protocol Dynamic Source Routing (DSR)

Queue type Priority Queue
Packet size 300 bytes
MAC type 802.11
Antenna Omni-directional antenna

Initial energy 100J
Idle node energy 0.001J/s

Power required for transmission 1.35J
Power required for reception 1.7J

Power consumed by sleep nodes 0.001J
Distance between Source and Destination 669.9Km
Energy consumed by the highest antenna 1.5 J/s

Propagation model Two-way ground

Figure 5: PDR comparison of the proposed system with existing techniques

and the attackers between the source and destination are discovered. To avoid packet loss, it stores
the attacker’s location in memory after it has been discovered.
The proposed fog-based method produces improved PDR outcomes in the vast majority of instances.
However, if an attacker is present, the packet delivery ratio is lowered. However, it outperforms other
key management systems currently in use. Figure 5 shows the PDR of the proposed fog-based system
in comparison to existing approaches. The suggested MRV-BSKG has a packet delivery ratio of 96
percent when compared to other current approaches such as LDPA (88 percent), HKM (90 percent),
and KMS-CL (94.45 percent).

4.2 Packet Drop

Packet drop mainly occurs in the wireless network during data transmission and network con-
gestion. It exhibits the reliability of the communication process in wireless networks. It is defined
as the number of packets sent from the source PS to the number of packets dropped (PND) during
transmission.



https://doi.org/10.15837/ijccc.2022.6.4363 10

Figure 6: Comparison of packet loss

The proposed MRV method’s packet loss is compared to current approaches. The findings are
shown in Figure 6 above. The efficiency of the proposed MRV-BSKG approach is determined by
altering the number of nodes and assessing the time associated with packet loss. The findings reveal
that the proposed MRV-BSKG approach detects packet drops in 3.75 seconds owing to attackers and
mobile stations, which is better than other methods currently in use. KMS-CL is 4.85 with ms, LPDA
is 4.17 with sec, HKM is 4.38 with sec, and LPDA is 4.17 with sec.

4.3 Throughput

It is defined as the amount of information sent from source to destination within a specific amount
of time. The following equation can compute the throughput of the wireless network. PR indicates
the packet received and PSindicate the packet size received in a time (T) sec.

Throughput(bps) = PR ∗PS/T (13)

The comparison graph demonstrates that the suggested method outperforms other current strategies in
terms of throughput. At 20ms, the suggested Fog-based system enhances the throughput comparison
given in figure 7; this demonstrates that the proposed technique has a higher throughput ratio than
previous approaches.



https://doi.org/10.15837/ijccc.2022.6.4363 11

Figure 7: Throughput comparison of proposed Fog based system with other existing techniques

Figure 8: Security Level Analysis

Figure 7 depicts the throughput curve of the proposed Fog-based system using the MRV-BSKG
key management approach. This graph indicates that as time passes, the system’s throughput grows.
The suggested Fog-based system’s throughput is compared to existing key management systems.

4.4 Computation Time

The time required for computing the shortest path between the source and destination in the
proposed Fog based system is shown in Figure 8.

Figure 8shows that the MRV reduces the latency by 23.14%, 48.82% and 43.79% compared with
conventional shortest path computation algorithms such as LPDA, HKM.

4.5 Security Analysis

Fog-based system’s major concern is network security because there are some adversaries to hack
the data during data transmission from source to destination. To mitigate these security and privacy
issues highly secured BSKG key generation-based fog system is designed that can provide 98% secured



https://doi.org/10.15837/ijccc.2022.6.4363 12

Figure 9: The computation time of shortest path algorithms

Figure 10: Data Decryption Time

data. Figure9 shows the results; it is observed that the proposed BSKG based security level is 98%
increases the security by when compared with the existing key management techniques such as ABE
is 90%, CP-ABE is 92%, and LDPA is 93

4.6 Analysis of Error Rate

It can be seen that the Key Bit Error Rate increases as the data size increases due to the decreased
subspace distance between potential precedes. The above figure 10 shows that the analysis of error
rate compared with proposed MRV-BSKG is 36% and the previous ABE is 67%, CP-ABE is 63%,
LDPA is 55%, HKM is 49%, KMS-CL is 41%.

4.7 Analysis of Encryption Time

Analysis of encryption time that indicates the speed of encryption. The experiments for comparing
the performance of the different previous security algorithm and proposed algorithm.

Figure 11 shows that encryption time for uploading the data and then it indicates how long it took
to convert the original data. The proposed MRV-BSKG is 1.87 with ms and the existing methods



https://doi.org/10.15837/ijccc.2022.6.4363 13

ABE with 2.9 ms, CP-ABE with 3.1 ms, LDPA with 2.6 ms, and HKM with 2.4 ms.

4.8 Decryption Time

The proposed MRV-BSKG algorithm providing a separate key to each user requesting the authen-
tication of necessary data posted by the data owner.

Indicates how long it took to convert the encrypted data and retrieve the original data. The
proposed MRV-BSKG algorithm it should be used once to decrypt the required information. The
above figure 12 shows the comparison of the proposed and previous methods. Hence the proposed
MRV-BSKG to give better decryption time

5 Conclusion
Fog- Cloud Computing is considered a good partner; it is an extended cloud service by end users

.The fundamental here is to interface and change the safety efforts and apply them as per the necessi-
ties of the Fog platform. The secured key management efforts have gone through testing, and utilizing
them can guarantee that any Fog computing fulfills vital mechanical security norms. Various security
issues may arise in the design and implementation of the technology. In this paper, the Fog-based
system is proposed with a hybrid BSKG method formed by combining two key generation approaches,
LKG and LBK. This method achieved high-level security against node capture attacks. The LKG
method is used for both encryption and decryption, and LBK is used to generate key in fog nodes.
The generated key is difficult to express because of its high dispersion nature. Analysis shows that the
proposed method achieves all security requirements like location privacy,anonymity,data integrity. Fu-
ture work may lead to the development of knowledge-based supplementary and aided systems, which
can provide decision support services to developers in designing a safe and efficient fog infrastructure.

Declaration of Competing Interests

The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

Conflict of interest

The authors declare no conflict of interest.

Ethics Approval and Consent to Participate

No participation of humans takes place in this implementation process

Human and Animal Rights

No violation of Human and Animal Rights is involved.

Funding

No funding is involved in this work.

Authorship contributions

There is no authorship contribution.

Acknowledgement

There is no acknowledgement involved in this work



https://doi.org/10.15837/ijccc.2022.6.4363 14

References
[1] Haina Zheng, KeXiong, Pingyi Fan, ZhangduiZhong and Khaled Ben Letaief, 2019.“Fog-Assisted

Multi-User SWIPT Networks: Local Computing or Offloading”, IEEE Internet of Things Jour-
nal, pp. 5246-5264.

[2] Andrea Tassi, IoannisMavromatis, Robert Piechockiy and Andrew Nix, 2019.“Agile Data Of-
floading over Novel Fog Computing Infrastructure for CAVs”, Networking and Internet Archi-
tecture.

[3] Mahesh U. Shankarwar and Ambika V. Pawar, 2016.“Security and Privacy in Cloud Computing:
A Survey”, Springer.

[4] MadjidMerabti, Bashar Alohali and KashifKifayat,2015. “A New Key Management Scheme
Based on Smart Grid Requirements”, Advances in Information Science and Computer Engi-
neering, pp. 436-444.

[5] W. Wang and Z. Lu, 2013 “Cyber security in the smart grid: Survey and challenges,” Computer
Networks, vol. 57, no. 5, pp. 1344–1371.

[6] M. Benmalek and Y. Challal, 2016 “MK-AMI: Efficient multi-group key management scheme
for secure communications in AMI systems,” in Proc. of IEEE Wireless Communications and
Networking Conference (IEEE WCNC), pp. 1–6.

[7] K. K. R. Choo, O. F. Rana, and M. Rajarajan. 2017“Cloud Security Engineering:Theory, Prac-
tice and Future Research.” IEEE Transactions on CloudComputing, 5(3):372–374.

[8] .O. A. Osanaiye, S. Chen, Z. Yan, R. Lu, K. K. R. Choo, and M. E.Dlodlo. 2017 From “Cloud
to Fog Computing: A Review and a ConceptualLive VM Migration Framework”. IEEE Access,
5(1):8284–8300.

[9] AshkanYousefpour et al, 2017 “All one needs to know about fog computing and related edge
computing paradigms: A complete survey”, Elsevier, Journal of Systems Architecture, Vol. 89,
pp. 289-330.

[10] Shanhe Yi, Zhengrui Qin, and Qun Li, 2015 “Security and Privacy Issues of Fog Computing: A
Survey”, Conference on wireless algorithms, systems, Springer.

[11] P. Zhang, T. Zhuo, W. Huang, K. Chen, M. Kankanhalli, Online object tracking based on CNN
with spatial-temporal saliency guided sampling, Neurocomputing 257 (2017) 115–127.

[12] J. Zhang, K.A. Ehinger, H. Wei, K. Zhang, J. Yang, A novel graph-based optimization framework
for salient object detection, PatternRecognit. 64 (1) (2017) 39–50.

[13] H. Chen, Y. Li, D. Su, Multi-modal fusion network with multi-scale multi– path and cross-modal
interactions for RGB-D salient object detection, Pattern Recognit. 1 (1) (2018).1–1.

[14] E. Macaluso, C.D. Frith, J. Driver, Directing attention to locations and to sensory modalities:
multiple levels of selective processing revealed with PET, Cerebral Cortex 12 (4) (2002) 357–368.

[15] T.S. Lee, D. Mumford, Hierarchical bayesian inference in the visual cortex, JOSAA 20 (7) (2003)
1434–1448.

[16] Q. Yan, L. Xu, J. Shi, J. Jia, Hierarchical saliency detection, in: IEEE Conference on Computer
Vision and Pattern Recognition, 2013, pp. 1155–1162.

[17] . Achanta, R., Hemami, S., Estrada, F., Susstrunk, S.: Frequency-tuned salient region detection.
In: IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2009, pp. 1597–1604.
IEEE (2009)



https://doi.org/10.15837/ijccc.2022.6.4363 15

[18] Cheng, M.M., Zhang, G.X., Mitra, N.J., Huang, X., Hu, S.M.: Global contrast based salient re-
gion detection. In: 2011 IEEE Conference on Computer Vision and Pattern Recognition (CVPR),
pp. 409–416. IEEE (2011)

[19] Cui, X., Liu, Q., Metaxas, D.: Temporal spectral residual: fast motion saliency detection. In:
Proceedings of the ACM International Conference on Multimedia (2009).

[20] B. X. Nie, P. Wei, and S.-C. Zhu, “Monocular 3D human pose estimation by predicting depth
on joints.” in IEEE International Conference on Computer Vision, 2017

[21] D. Zhang, J. Han, C. Li, J. Wang, and X. Li, “Detection of co-salient objects by looking deep
and wide”, International Journal of Computer Vision, vol. 120, no. 2, pp. 215–232, 2016.

[22] X. Dong et al., “Occlusion-aware real-time object tracking,” IEEE Trans. Multimedia, vol. 19,
no. 4, pp. 763–771, Apr. 2017.

[23] X. Dong, J. Shen, L. Shao, and L. Van Gool, “Sub-Markov random walk for image segmentation,”
IEEE Trans. Image Process., vol. 25, no. 2, pp. 516–527, Feb. 2016.

[24] J. Shen et al., “Real-time superpixel segmentation by DBSCAN clustering algorithm”, IEEE
Trans. Image Process., vol. 25, no. 12, pp. 5933–5942, Dec. 2016.

[25] Y. Yuan, C. Li, J. Kim, W. Cai, D.D. Feng, Dense and sparse labeling with multidimensional fea-
tures for saliency detection, IEEE Trans. Circuits Syst. Video Technol. 28 (5) (2018) 1130–1143.

[26] W. Wang, J. Shen, F. Guo, M.-M. Cheng, A. Borji, Revisiting video saliency: a large-scale
benchmark and a new model, in: IEEE Conference on Computer Vision and Pattern Recognition,
2018, pp. 4894–4903.

[27] Li Q., Chen S., Zhang B. (2012) Predictive Video Saliency Detection. In: Liu CL., Zhang C.,
Wang L. (eds) Pattern Recognition. CCPR 2012. Communications in Computer and Information
Science, vol 321. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-33506-8_23.

[28] Wang, Wenguan et al. “Deep Learning For Video Saliency Detection.” ArXiv abs/1702. 00871
(2017): n. pag.

[29] F. Guo et al., "Video Saliency Detection Using Object Proposals," in IEEE Transactions on
Cybernetics, vol. 48, no. 11, pp. 3159-3170, Nov. 2018, doi: 10.1109/TCYB.2017.2761361.

[30] Karthik, A., MazherIqbal, J.L. Efficient Speech Enhancement Using Recurrent Convolution
Encoder and Decoder. Wireless Pers Commun 119, 1959–1973 (2021).

[31] Yuming Fang, Xiaoqiang Zhang, Feiniu Yuan, NevrezImamoglu, Haiwen Liu, Video saliency
detection by gestalt theory, Pattern Recognition, Volume 96,2019,106987, ISSN 0031-3203.

[32] https://docs.microsoft.com/en-us/cpp/build/reference/clr common language runtime compila-
tion? View = msvc-160

[33] https://docs.microsoft.com/en-us/cpp/dotnet/walkthrough-compiling-a-cpp-program-that-
targets-the-clr-in-visual-studio?view=msvc-160

[34] https://en.wikipedia.org/wiki/Common_Language_Runtime

[35] https://www.red-gate.com/simple-talk/dotnet/net-development/creating-ccli-wrapper/

[36] Wang, Bofei et al. “Object-based Spatial Similarity for Semi-supervised Video Object Segmen-
tation.” (2019).

[37] Li F., Kim T., Humayun A., Tsai D., Rehg J. M.,“Video Segmentation byTracking Many Figure-
Ground Segments” In:IEEE International Conference onComputer Vision (ICCV), 2013.



https://doi.org/10.15837/ijccc.2022.6.4363 16

Copyright ©2022 by the authors. Licensee Agora University, Oradea, Romania.
This is an open access article distributed under the terms and conditions of the Creative Commons
Attribution-NonCommercial 4.0 International License.
Journal’s webpage: http://univagora.ro/jour/index.php/ijccc/

This journal is a member of, and subscribes to the principles of,
the Committee on Publication Ethics (COPE).

https://publicationethics.org/members/international-journal-computers-communications-and-control

Cite this paper as:
Arul Sindhia.P; Bharathi.R; (2022). Bi-Level Minimal Resource Protected Key Generation Framework
For Fog Computing Applications, International Journal of Computers Communications & Control,
17(6), 4363, 2022.

https://doi.org/10.15837/ijccc.2022.6.4363


	Introduction
	Related work
	Implementation of the Proposed System 
	Proposed Fog based System
	Bi-level Secured Key Generation (BSKG)
	Lagrange’s Key Generation (LKG))
	Location-based Key Generation (LBK)

	Minimal Resource Viterbi

	Results and Discussions
	Packet Delivery Ratio (PDR)
	Packet Drop
	Throughput
	Computation Time
	Security Analysis
	Analysis of Error Rate
	Analysis of Encryption Time 
	Decryption Time 

	Conclusion