The importance of virtualized infrastructures and cloud
computing is currently increasing rapidly. Virtual
infrastructures allow servers, networks, and storage to
be virtualized and shared between different users.
Cloud computing generalizes and automates this
approach such that users of a d ata center can request
virtually any number of machines, networks, and
storage while provisioning and scaling is fast and
managed transparently by the provider [12].
The increasing complexity and multitenancy of such
virtualized infrastructures can cause severe security
problems due to possible misconfigurations, e.g. two
different users have access to the same storage, and the
abstraction of cloud computing hinders the verification
of policy compliance. An automated mechanism is
required to handle these scenarios and IBM built a
prototype for retrieving the configuration of virtual
systems and performing certain security analysis on
them[12].
Cloud computing has gained remarkable popularity in
the recent years by a w ide spectrum of consumers,
ranging from small start-ups to governments [12].
However, its benefits in terms of flexibility, scalability,
and low upfront investments, are shadowed by security
challenges which inhibit its adoption. In particular,
these highly flexible but complex cloud computing
environments are prone to misconfigurations leading to
security incidents, e.g., erroneous exposure of services
due to faulty network security configurations.
In recent years, Cloud Computing has gained
remarkable popularity due to the economical and
technical benefits provided by this new way of
delivering computing resources, and the pervasive
availability of high-speed networks. Businesses can
offload their IT infrastructure into the cloud and benefit
from the rapid provisioning and scalability. This allows
an on-demand growth of IT resources in addition to a
pay-as-you-go pricing scheme, which does not require
a high up-front capital investment. These benefits are
in particular attractive to small businesses, like start-
ups, who often have traffic spikes or a steep growth
rate, and who prefer to avoid intensive up-front capital
investment in their IT infrastructure [12]. However,
cloud computing is not limited to such small business.
The US government, one of the largest consumers of
information technology, is initiating a move of parts of
its IT infrastructure into the cloud, in order to reduce
costs and gain productivity. These general principles of
cloud computing can be implemented on d ifferent
abstraction levels. While Infrastructure as a S ervice,
such as Amazon EC2, provides virtual m achines,
storage, and networks, higher abstractions include
Platform as a Service as well as Software as a Service
that provide the actual web-based applications to end-
users [12].
Despite its benefits, Cloud Computing also induces
unique challenges in terms of security. Multi-tenancy
requires proper isolation of users, the abstraction of the
cloud hinders compliance verification of the underlying
1. Introduction
Optimizing Cloud Security by Applying New Innovative Filter
MEHDI DARBANDI
Department of Electrical Engineering and Computer Science at Iran University of Science and
Technology, IRAN
Abstract: In this paper, at first authors discuss about principles of Cloud Computing and basic concepts
of this new and ground-breaking technology. After this brief introduction, they study more about
influences of this technology on different industries and products; they specially focus on Amazon
products and their future decisions. After that, in the second section, they present new and innovative
filter which can be used as an estimator in cloud platforms. Authors claim that if service providers use
such innovative algorithms and equip their gateways and routers with such algorithm they are able to
estimate and predict about lots of critical criteria’s. For example, they can estimate the presence of
hackers or race them when they’re inside of the network and eliminate them finally or even they can be
able to estimate and predict the amount of resources which are need in specific time to prevent from
wasting of resources or sudden crashing. Authors of this paper proof their new algorithm by
mathematical equations and several simulations at the end of their papers.
Keywords: Cloud platforms security, new innovative Kalman Filter, hacker tracing.
Received: June 25, 2021. Revised: November 29, 2021. Accepted: December 15, 2021. Published: January 3, 2022.
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
1
Volume 21, 2022
architecture, and the sheer complexity of such a system
implies a high probability of misconfigurations
endangering the overall security. While the benefits of
cloud computing are clear and end-users demand such
services, security is a major inhibitor of cloud
computing adoption on all levels of abstraction. In
numerous studies the security related problems have
been pointed out. One of the top risks exposed in the
study is the failure of isolation in the cloud computing
environment [12].
Cloud computing environments are becoming
increasingly complex, more tenants are sharing the
same physical resources, and the flexibility and
possibility of programmatic configurations can lead to
unforeseen misconfigurations. For example, network-
based storage volumes can be flexibly attached to
virtual machines, and potentially a volume will be
attached to a wrong virtual machine risking the
exposure of sensitive data on t hat volume. Network
security is also flexibly managed through a
programmatic interface, which could lead to problems
resulting in network services exposed wrongly to the
public and opening not properly secured services to
other peers [12]. Administrators of such virtual
infrastructures must be able to easily understand the
complex deployments and ensure that proper security
is given. The dynamic and agility of such
environments also provides a challenge in ensuring the
security over its entire lifetime due to their constant
changes [12].
In order to successfully address the problem of
configuration complexity and potential
misconfigurations in cloud computing environments,
we narrowed down the problem domain to a specific
case of multi-tier applications deployed in
infrastructure clouds using a specific cloud provider as
an example case. We will study existing literature in
the broad domain of virtual machine security, which
plays a f undamental part in the security of
infrastructure clouds, and network security analysis
with a focus on vulnerability assessment and reach-
ability [12]. B ased on the insights and inspirations
obtained by performing the literature review, we will
propose a novel approach in assessing the security of a
multi-tier application deployed on the Amazon
infrastructure cloud. By implementing our approach
and then evaluating it regarding practicality and
scalability, we will determine the practical usefulness
for detecting misconfigurations even in large-scale
deployments. The evaluation is performed both
theoretical and practical. The theoretical evaluation is
conducted by assuming complex configuration
scenarios and analyzes the algorithm run-time using an
ideal computer. The practical evaluation is performed
using the implementation on a sample multi-tier
application deployed on Amazon EC2 [12].
The main contribution of this paper is a novel approach
in the security evaluation of multi-tier virtual
infrastructures, inspired by vulnerability assessment
approaches for traditional computing environments and
applied for the case of the Amazon infrastructure
cloud. The security evaluation consists of an automated
security audit process of the currently deployed
configuration with regard to a given policy specifying
the desired state of the configuration, and an abstract
framework for evaluating the security impact of
configuration changes. Besides the main contribution
stated above, multiple minor contributions can be
pointed out [12]. A comprehensive description of the
underlying architecture of the Amazon infrastructure
cloud is presented, which was publicly only available
in incomplete and fragmented form. We provide a
comparison of two methods for deploying multi-tier
virtual infrastructures on Amazon with regard to the
provided isolation levels [12]. Finally, a data model for
representing the configuration of Amazon deployments
is presented and integrated into a larger data model
capable of representing configurations of different
virtualization systems [12].
Cloud computing is a broad term combining several
different types of service offerings. In general we
distinguish between Software, Platform, and
Infrastructure as a se rvice, which are offered by the
cloud provider [12]. The main focus of this paper lies
on Infrastructure as a Service, also called
Infrastructure Clouds, but for comparison reasons the
other types of offerings are also briefly presented [12].
Storage can be provided in different ways varying
among the multiple IaaS providers. Four different
forms can be identified in the currently available
providers: NAS-like, SAN-like, API-based data
objects, and Virtual Machine storage. A virtual
machine has typically a f ixed-size data storage
available, which is equivalent to a hard disk in a
regular desktop or server computer. In some cases this
type of storage is only intended to be used for
temporary data and is itself non-persistent, i.e., after
the machine terminates the data is lost [12]. NAS-like
storage, like GoGrid Cloud Storage, is accessible from
the VMs on a file-based level using standard protocols
like CIFS. Amazon Elastic Block Store (EBS) is a
SAN-like storage type, which appears to the VM as an
additional block-device. An EBS volume can be
attached to different VMs, but not to multiple VMs
simultaneously, and the size can be adjusted presuming
the file system on the block-device is resizable as well.
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
2
Volume 21, 2022
The last type of storage is accessible through an API
and holds data objects up to a specific size, e.g., in the
range of several gigabytes. This is a very scalable kind
of storage, i.e., one can store an arbitrary amount of
objects, and also provides the possibility of distributing
these objects using a Content D istribution Network
offered by the provider. Examples of this kind of
storage are Amazon Simple Storage Service (S3) and
RackSpace CloudFiles [12]
1. Statement of the Problem:
The problem of interest is described by the discretized
equation set [13]:
x
kkkkkk
WUBXAX ++=
+1
(1)
u
kkkk
WUCU +=
+1
(2)
kkkk
VXHZ +=
(3)
Where
n
k
RX ∈
is the system state,
m
k
RU ∈
and
p
kRZ ∈ are the input and the measurement vectors,
respectively. Matrices
k
A
,
k
B
,
k
C
and k
H
are assumed
to be known functions of the time interval
k
and are of
appropriate dimensions. Matrix
k
C
is assumed
nonsingular. The process noises
x
k
W,
u
k
W
and the
measurement noise k
V
are zero-mean white Gaussian
sequences with the following covariance’s:
kl
x
k
x
l
x
k
QWWE
δ
=])([
'
,
kl
xu
k
u
l
x
k
QWWE
δ
=])([
'
,
kl
u
k
u
l
u
k
QWWE
δ
=])([
'
klklk
RVVE
δ
=][
'
,
0][
'
=
l
x
k
VWE
and
0][
'
=
l
u
k
VWE
, where
'
denotes transpose and
kl
δ
denotes the Kronecker delta function. The initial states
0
X
and
0
U
are assumed to be uncorrelated with the
sequences
x
k
W
,
u
k
W
and k
V
. The initial conditions are
assumed to be Gaussian random variables
with
00 ˆ
][ XXE =
,
x
PXXE
0
'
00
][ =
,
00 ˆ
][ UUE =
,
u
PUUE
0
'
00
][ =
,
xu
PUXE
0
'
00
][ =
.
Treating k
X
and
k
U
as the augmented system state,
the AUSKE is described by [13]:
)( |1111|11|1
Aug
kk
Aug
kk
Aug
k
Aug
kk
Aug
kk XHZKXX +++++++ −+=
(4)
Aug
kk
Aug
k
Aug
kk
XAX
||1
=
+
(5)
1'
1|11
'
1|11 ])([)( −
++++++ += k
Aug
kkk
Aug
k
Aug
kkk
Aug
kRHPHHPK
(6)
k
Aug
kkk
Aug
kkk
QAPAP +=
+
'
||1
)(
(7)
kk
Aug
k
Aug
kkk
PHKIP
|1111|1
)(
+++++
−=
(8)
Where
=
k
k
Aug
kU
X
X
,
=u
k
x
k
Aug
kK
K
K
,
=
u
k
xu
k
xu
k
x
k
k
PP
PP
P
'
)(
,
=
×knm
kk
Aug
k
C
BA
A0
,
'
0
=
×mp
k
Aug
k
H
H
,
=
u
k
xu
k
xu
k
x
k
k
QQ
QQ
Q
'
)(
Where the superscript ‘Aug’ denotes the augmented
system state,
I
denotes the identity matrix of any
dimension and
nm×
0
is a
nm ×
zero matrix. It is clear
from (4)-(8) that the computational cost of the AUSKE
increases with the augmented state dimension. The
OPSKE formulation is based on t he following
equations [13]:
)
ˆ
(
ˆˆ
|1111|11|1 kkkkkkkkk
XHZKXX
+++++++
−+=
(9)
kkkkk
XAX
||1
ˆˆ =
+
(10)
1'
1|11
'
1|11 ])([ −
++++++ += kk
x
kkkk
x
kkk RHPHHPK
(11)
x
kk
x
kkk
x
kk
QAPAP +=
+
'
||1
)(
(12)
x
kkkk
x
kk
PHKIP
|1111|1
)(
+++++
−=
(13)
1111 ++++
−=
kkkk
M]HKI[N
(14)
]
ˆ
~
[
ˆˆ
|11111|11|1 kkkkk
u
kkkkk
UMHZKUU
++++++++
−+=
(15)
kkkkk UCU ˆˆ 1=
+
(16)
1
|1
'
1
'
1|111
'
1
'
1|11
]3[2
−
++++++++++
+×=
z
kkkk
u
kkkkkk
u
kk
u
k
PHMPMHHMPK
(17)
u
kkkk
u
k
u
kkk
u
kk
u
k
z
kk
u
k
u
kkk
u
kkkk
u
k
u
kk
u
kk
PMHKKHMPKPK
KHMPMHKPP
|11111
'
1
'
1|11|11
1
'
1
'
1|1111|11|1
2)(2)(
)(3
+++++++++++
++++++++++
−
′
−
′
+
′
+=
(18)
u
kk
u
kkk
u
kk
QCPCP +=
+
'
||1
(19)
1
'
1|11|1 +++++
+=
kk
x
kkk
z
kk
RHPHP
(20)
u
kkkk
zu
kk PMHP |111|1 ++++ =
(21)
11|1|1
ˆ
ˆ++++ += kkkkkk UMXX
111|11|1
ˆ
ˆ++++++ += kkkkkk UNXX
(22)
1
001
1
1,....3,2,][
−
−
+
=
=+=
CBM
kCBMAM kkkkk
(23)
1111
][
++++
−=
kkkk
MHKIN
(24)
2. Performance Evaluations [13]:
To demonstrate the computational advantage of the
OPSKE over the AUSKE, the number of arithmetic
operations are considered, i.e., multiplications and
summations. The arithmetic operations of a standard
Kalman estimator with state dimension
n
and
measurement dimension
p
, are listed in Table 1. It is
clear from the equations (4)-(8) and Table 1, that the
arithmetic operations required for the AUSKE which
has state dimension
mn +
and measurement dimension
2. Performance Comparison of Two
Stage Kalman Filtering Technique for
Surveillance Permeating Tracking in
Cloud Computing [17]:
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
3
Volume 21, 2022
p
, are
),( pmnM +
for multiplications and
),( pmnS +
for
summations. Table 2 shows the arithmetic operations
of the input estimation and the auxiliary matrices
needed by the OPSKE which has state dimension
n
,
measurement dimension
p
and input vector dimension
m
. Note that the number of the arithmetic operations
of the AUSKE increases with the augmented state
dimension, which makes the algorithm computationally
inefficient. In contrast, the OPSKE based on the two-
stage decoupling technique required fewer
computations. The efficiency of the OPSKE is due to
order reduction, i.e., implementing two less order
n
and
m
partitioned filters. This enables the proposed
algorithm to have much better computational efficiency
than the AUSKE. So, the arithmetic operations
required (AOR) for the AUSKE are [13]:
)]()()(2)(2)(3[
])(2)()(2)(2)(3[
),(),()(
23223
23223
mnmnppmnpmnmn
pmnmnppmnpmnmn
pmnSpmnMAUSKEAOR
+−+−+++++++
++++++++++=
+++=
(25)
The arithmetic operations required for the input
estimation and auxiliary matrices, by the OPSKE as
shown in Table 2 and using equations (15)-(24) are
]222
422[
]224
2223[
]223[
]2223[
),,(),,(),(),(
)(
2222
33222
2223
23222
23223
23223
nmpnmmnnpnppn
mpmppmmmmp
nmpnmmnnmpnm
ppmppmmmp
nnpnppnn
npnpnppnn
pmnSpmnMpnSpnM
OPSKEAOR
OPOP
+++−+−+
++++−−−+
++++++
++++++
−−++++
+++++=
+++=
(26)
Using (25) and (26), the operational savings, denoted
by
OPSKE
AUSKE
OS
, of the OPSKE as compared to the
AUSKE are [13]:
nmpmnnpp
nmppnnmmnm
pmnSpmnMpnS
pnMpmnSpmnM
OPSKEAORAUSKEAOROS
OPOP
OPSKE
AUSKE
2222
6417152
),,(),,(),(
),(),(),(
)()(
223
2223
−−−++−
+−++−=
−−−
−+++
=−=
(27)
And the operational savings of the OTSKE over the
AUSKE are:
nmpmmnmpnmmn
mOTSKEAORAUSKEAOROS OTSKE
AUSKE
2241212
4)()(
3222
3
−−−+++
+−=−=
(28)
Therefore, using (27) and (28) the operational savings
of the OPSKE over the OTSKE are [13]:
22322
23
22245
32)()(
pmmnnppnmppnnm
mnmOPSKEAOROTSKEAOROSOPSKE
OTSKE
−+−++−+−+
+=−=
(29)
It is clear from (27) and (29) that for
npm ≤and
, the
proposed scheme has computational advantage over the
AUSKE and it is comparable to the OTSKE. The
operational savings discussed here will be tested as an
example in the simulation results section. To measure
the relative operational savings of the OPSKE with
respect to the arithmetic operation required by the
AUSKE (
)(AUSKEAOR
), the percentage of the
operational savings defined as below:
100
)( ×= AUSKEAOR
OS
POS
OPSKE
AUSKE
OPSKE
AUSKE
(30)
Using (27), (29) and (30), the operational savings and
the percentage of the operational savings, of the
OPSKE comparing to the OTSKE and the AUSKE for
different values of
n
,
m
and
p
are shown in Table 3.
It can be inferred from Table 3 t hat the OPSKE has
better overall performance than the AUSKE (averaged
32%) and the OTSKE (averaged 7.3%) [13].
Table 1:Standard Kalman Estimator Arithmetic Operation Requirements [13]
Variable
Number of Multiplications,
)p,n(M
Number of summations,
)p,n(S
1
11 ++ k|k
X
np2
np2
2
k|K
X1+
2
n
nn −
2
3
x
k
K1+
322
2pnppn ++
nppnppn 22 322 −++
4
x
k|K
P1+
3
2n
23
2nn −
5
x
k|K
P11 ++
pnn 23 +
223
npnn −+
Totals
npnpnppnn 2223
23223
+++++
nnpnppnn −−+++
23223
223
Table 2:Input Estimation and Auxiliary Matrices Arithmetic Operation Requirements for the OPSKE [13]
Variable
Number of Multiplications
)p,m,n(M OP
Number of summations
)p,m,n(S
OP
1
11 ++ k|k
U
mp2
mp2
2
k|K
U
1+
2
m
mm −
2
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
4
Volume 21, 2022
3
u
k
K1+
mpppmppm ++++
2322
2
mppmppm 22 322 −++
4
u
k|K
P1+
3
2m
23
2mm −
5
u
k|K
P11 ++
223 mpmm ++
223
mpmm −+
6
z
k|k
P1+
pn2
2
22 22 pnppn +−
7
k|k
X
ˆ
1+
mn
mn
8
11 ++ k|k
X
ˆ
mn
nmn −
9
1+k
M
232
nmmmn ++
nmnmmmn −++ 232
10
1+k
N
mn2
nmmn −
2
11
11 ++ kk
MH
nmp
mpnmp −
Totals
nmpnmmnnmpnm
ppmppmmmp
++++++
+++++
2223
23222
224
2223
nmpnmmnnpnppn
mpmppmmmmp
+++−+−+
++++−−−
2222
33222
222
422
Table 3:the Operational Savings and the Percentage of the Operational Savings of the OPSKE
Compared to the AUSKE and the OTSKE [13]
The state vector
dimensions
OPSKE
AUSKE
OS
OPSKE
AUSKE
POS
(%)
OPSKE
OTSKE
OS
OPSKE
OTSKE
POS
(%)
244 === p,m,n
1340
35.7
592
15.7
224 === p,m,n
578
33.7
102
5.9
124 === p,m,n
553
37.5
155
10.5
114 === p,m,n
242
27.5
23
2.6
334 === p,m,n
978
32.7
247
8.2
2210 === p,m,n
2954
25.1
132
1.12
Average
≅
1107
32.0
≅
208
7.3
3. Simulation Results:
To evaluate the proposed algorithm, an example of
maneuvering target tracking problem which turns, in
two-dimensional space is simulated such as
permeating a hacker into a very important network or
databases. In this simulation example, the
performance of the OPSKE for the maneuvering
target tracking has been compared with the traditional
works that done in this concept, as an example of the
AUSKE method. As mentioned before in the
augmented state method the state vector includes the
input vector i.e., acceleration and jerk parameter in
maneuvering target tracking problem. The sampling
interval is T=0.01 (sec) and target maneuver is applied
at 9th second (900th sample). The initial conditions
are selected similar for the AUSKE as w ell as the
OPSKE. The state vectors are
[ ]
'
y
kk
x
kkk vyvxX =
,
[ ]
'
y
k
y
k
x
k
x
kk jujuU =
,
[ ]
'
y
k
y
k
x
k
x
k
y
kk
x
kk
Aug
k
jujuvyvxX =
Where
k
x
,
x
k
v
,
x
k
u
and
x
k
j
denote the position,
velocity, acceleration and jerk of the target along the
x
axis, respectively. We consider the target initial
conditions for the state and the acceleration vectors as
below [13]:
[ ]
' s/m m s/m m X 251250802165
0
−=
,
[ ]
' sec/g g sec/g g U 0000
0
=
[ ]
' sec/ 0 0sec/ 0 0/ 25 1250/ 80 2165
0
ggggsmmsmmX
Aug
−=
The target begins to maneuver as
[ ]
' sec/ 4.00sec/ 7.00
900 ggggU −=
for
sec)( 90(sec) 9 ≤≤ t
.
The system matrices are given by
=
1000
100
0010
001
T
T
Ak
,
=
2/00
6/2/00
002/
006/2/
2
32
2
32
TT
TT
TT
TT
B
k
,
=
1000
100
0010
001
T
T
C
k
,
'
00
10
00
01
=
k
H
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
5
Volume 21, 2022
=
TT
TT
TT
TT
Q
j
u
k
2/00
2/3/00
002/
002/3/
2
2
23
2
23
ασ
,
=
20/72/00
72/252/00
0020/72/
0072/252/
2
56
67
56
67
TT
TT
TT
TT
Q
j
x
k
ασ
=
6/8/00
24/30/00
006/8/
0024/30/
2
34
45
34
45
TT
TT
TT
TT
Q
j
xu
k
ασ
,
440 10 ×
=IP x
,
44
1.0 ×
=IPu
o
,
440 ×
=IP
xu
,
'
42
0
=
×
k
Aug
k
H
H
=
×k
kk
Aug
kC
BA
A
44
0
,
=
u
k
xu
k
xu
k
x
k
k
QQ
QQ
Q
'
)(
,
=u
k
xu
k
xu
k
x
k
kPP
PP
P'
)(
.
Where
)(09.0
3−
=ms
j
σ
the variance of the target is jerk
and
)(s 0123.0 -1
=
α
is the reciprocal of the jerk time
constant
ατ
/1=
. T he measurement standard
deviations of
x
and
y
target positions are:
)( 1010 m
x=
σ
,
)( 20 m
y
=
σ
. Thus, the measurement
covariance matrix is
=4000
01000
k
R
for both methods
[13]. The Root Mean Square Error (RMSE) index is
used for the results evaluation.
Fig. 1 shows the actual value and the estimation of
x
and
y
and RMS errors of
x
and
y
positions
estimations by the proposed OPSKE and the AUSKE.
Fig. 2 shows the actual value and the estimations of
yx vv ,
and the RMS errors of the
x
and
y
velocities
estimations by the proposed method compared with
the augmented method. The actual value and the
accelerations estimations in the
x
and
y
directions
and their corresponding averaged RMS errors can be
seen in Fig. 3.Fig. 4 displays the actual value and the
estimated jerk parameters are evaluated by the
OPSKE and the AUSKE methodologies [13].
10 15 20 25
-3
-2
-1
0
x 10
4
Time (sec)
x (m)
Atcual position
OPSKE method estimation
AUSKE method estimation
10 15 20 25
0
100
200
300
400
500
Time (sec)
Averaged RMSE of x (m)
OPSKE
AUSKE
10 15 20 25
0
0.5
1
1.5
2
x 10
4
Time (sec)
y (m)
Atcual position
OPSKE method estimation
AUSKE method estimation
10 15 20 25
0
50
100
150
200
250
300
Time (sec)
Averaged RMSE of y (m)
OPSKE
AUSKE
Fig. 1. The actual value and the estimation of the x, y positions and RMS errors estimations by the OPSKE and
the AUSKE methods.
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
6
Volume 21, 2022
10 15 20 25
-5000
-4000
-3000
-2000
-1000
0
Time (sec)
v
x
(m/sec)
Atcual velocity
OPSKE method estimation
AUSKE method estimation
10 15 20 25
0
100
200
300
Time (sec)
Averaged RMSE of v
x
(m/sec)
OPSKE
AUSKE
10 15 20 25
0
1000
2000
3000
Time (sec)
v
y
(m/sec)
Atcual velocity
OPSKE method estimation
AUSKE method estimation
10 15 20 25
0
50
100
150
200
250
Time (sec)
Averaged RMSE of v
y
(m/sec)
OPSKE
AUSKE
Fig. 2. T he actual value and the estimation of
yx v ,v
and RMS errors of x and y velocities estimations by the
OPSKE and the AUSKE methods.
10 15 20 25
-15
-10
-5
0
Time (sec)
u
x
(m/sec
2
)
Atcual acceleration
OPSKE method estimation
AUSKE method estimation
10 15 20 25
0
0.5
1
1.5
2
Time (sec)
Averaged RMSE of u
x
(m/sec
2
)
OPSKE
AUSKE
10 15 20 25
0
2
4
6
Time (sec)
u
y
(m/sec
2
)
Atcual acceleration
OPSKE method estimation
AUSKE method estimation
10 15 20 25
0.2
0.4
0.6
0.8
1
1.2
Time (sec)
Averaged RMSE of u
y
(m/sec
2
)
OPSKE
AUSKE
Fig. 3. The actual value and the estimation of acceleration in x and y directions and corresponding RMS errors by
the proposed method compared with the augmented methods.
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
7
Volume 21, 2022
10 15 20 25
-0.6
-0.4
-0.2
0
time (sec)
j
x
(m/sec
3
)
Atcual jerk
OPSKE method estimation
AUSKE method estimation
10 15 20 25
0
0.1
0.2
0.3
0.4
0.5
0.6
time (sec)
Averaged RMSE of j
x
(m/sec
3
)
OPSKE
AUSKE
10 15 20 25
0.2
0.4
0.6
0.8
time (sec)
j
y
(m/sec
3
)
Atcual jerk
OPSKE method estimation
AUSKE method estimation
10 15 20 25
0
0.1
0.2
0.3
0.4
0.5
0.6
time (sec)
Averaged RMSE of j
y
(m/sec
3
)
OPSKE
AUSKE
Fig. 4. The actual value and the estimation of jerk parameters and RMS errors by the OPSKE method compared
with the AUSKE method.
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
8
Volume 21, 2022
It is clear that the performance of the proposed
OPSKE is as well as the results obtained by the
AUSKE in the maneuvering target tracking problem.
Note that in this example
4=n
,
4=m
and
2=p
, and
the operation savings for the OPSKE over the AUSKE
and the OTSKE as shown in Table 3 a re 1340 (or
35.7%) and 592 (or 15.7%), respectively.
In this paper, first of all, authors discuss about
different aspects of cloud computing and impacts of
this technology on different industries and societies,
they study about impacts and influences of this
technology with focus to the Amazon products. After
understanding this technology and attain more about
applications of this technology, they reveal their new
and novel algorithm, which is named as two-stage
Kalman filtering. Authors claim that by using such
algorithm we can estimate and predict about all
important factors that are dealing with using of such
networks.
References
[1] Mehdi Darbandi “Applying Kalman Filtering in
solving SSM estimation problem by the means of
EM algorithm with considering a practical
example”; published by the Journal of Computing
– Springer, 2012; USA.
[2] Mehdi Darbandi; “Comparison between
miscellaneous platforms that present for cloud
computing and accreting the security of these
platforms by new filter”; published by the Journal
of Computing – Springer, 2012; USA.
[3] Mehdi Darbandi; “New and novel technique in
designing electromagnetic filter for eliminating
EMI radiations and optimization performances”;
published by the Journal of Computing -
Springer, 2012; USA.
[4] Mehdi Darbandi; “Appraising the role of cloud
computing in daily life and presenting new
solutions for stabilization of this technology”;
published by the Journal of Computing -
Springer, 2012; USA.
[5] Mehdi Darbandi; “Cloud Computing make a
revolution in economy and Information
Technology”; published by the Journal of
Computing - Springer, 2012; USA.
[6] Mehdi Darbandi; “Considering the high impact of
gettering of silicon on fabrication of wafer
designing and optimize the designing with new
innovative solutions”; published by the Journal of
Computing – Springer, 2012; USA.
[7] Mehdi Darbandi; “Developing concept of
electromagnetic filter design by considering new
parameters and use of mathematical analysis”;
published by the Journal of Computing -
Springer, 2012; USA.
[8] Mehdi Darbandi; “Is the cloud computing real or
hype Affirmation momentous traits of this
technology by proffering maiden scenarios”;
published by the Journal of Computing –
Springer, 2012; USA.
[9] Mehdi Darbandi; “Measurement and collation
overriding traits of computer networks and
ascertainment consequential exclusivities of cloud
computing by the means of Bucy filtering”;
published by the Journal of Computing -
Springer, 2012; USA.
[10] Mehdi Darbandi; “Unabridged collation about
multifarious computing methods and outreaching
cloud computing based on innovative procedure”;
published by the Journal of Computing -
Springer, 2012; USA.
[11] Mehdi Darbandi; “Scrutiny about all security
standards in cloud computing and present new
novel standard for security of such networks”;
published by the Journal of Computing -
Springer, 2012; USA.
[12] MSc. Thesis of Sören Bleikertz; Norwegian
University of Science and Technology; June 2010.
[13] A. Karsaz, H. Khaloozade, M. Darbandi;
“Performance Comparison of the two-stage
Kalman filtering Techniques for Target Tracking”
Int. IEEE Conf. Harbin, China.
3. Conclusion
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the Creative
Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en_US
WSEAS TRANSACTIONS on COMMUNICATIONS
DOI: 10.37394/23204.2022.21.1
Mehdi Darbandi
E-ISSN: 2224-2864
9
Volume 21, 2022
