IEEE Transactions on Intelligent Transportation Systems,
Journal Year:
2023,
Volume and Issue:
24(4), P. 4158 - 4172
Published: Jan. 9, 2023
Falsification
detection
is
a
critical
advance
in
ensuring
that
real-time
information
about
vehicles
and
their
movement
states
certified
on
the
Internet
of
Vehicles
(IoV).
Thus,
detecting
nodes
are
propagating
inaccurate
requirement
for
successful
deployment
IoV
services
although
only
few
research
studies
have
been
carried
out
Basic
Safety
Message
(BSM)
falsification.
As
such,
this
paper
proposes
Randomized
Search
Optimization
Ensemble-based
Detection
Scheme
(RSO-FDS).
The
RSO
technique
was
used
to
construct
proposed
Random
Forest
(RF)
model.
evaluation
performed
three
different
datasets
developed
evaluate
falsification
IoV.
In
addition,
six
most
popular
supervised
learning
(SL)
algorithms
were
investigated
capability
RSO-FDS,
which
had
best
performance
across
all
datasets.
metrics
considered
computational
efficiency
terms
prediction
time,
validation
accuracy
overall
attack
classification,
precision,
recall,
F1
scores.
For
validation,
RSO-FDS
further
compared
with
results
from
recent
works.
Furthermore,
irrelevance
data
balancing
illustrated
real-life
scenarios.
result
shows
model
outperformed
state-of-the-art
implemented
work
related
IEEE Access,
Journal Year:
2024,
Volume and Issue:
12, P. 19216 - 19228
Published: Jan. 1, 2024
In
this
paper,
we
consider
and
analyze
the
performance
of
exploiting
simultaneously
wireless
information
power
transfer
(SWIPT)
for
coordinated-non-orthogonal
multiple
access
(NOMA)
systems
under
Nakagami-
m
fading
channels.
Specifically,
to
deal
with
limited
energy
issues
at
relay
unit
while
taking
into
reusing
resource
efficiency,
propose
use
SWIPT-based
splitting
mechanisms
in
conjunction
NOMA
approaches
by
means
proposing
two
separate
scenarios,
first
focus
is
on
forward-based
device-to-device
(D2D)
transmissions,
called
SWIPT-NOMA-D2D,
second
forwarding
signal
only,
SWIPT-NOMA-noD2D.
Evaluating
case
case,
derive
exact
approximate
closed-form
expressions
outage
probabilities
cell-center
cell-edge
users,
respectively.
Besides,
also
obtain
an
solution
ergodic
rate
message
transmission,
which
enables
us
effectively
characterize
system
spectrum
utilization.
The
correctness
our
analyses
tightness
approximation
developed
are
validated
through
Monte
Carlo
simulations
numerical
techniques.
Finally,
show
that
proposed
schemes
achieve
near
sum
cases
unconstrained
designs
their
effectiveness
improving
centre-cell
users.
IEEE Transactions on Parallel and Distributed Systems,
Journal Year:
2024,
Volume and Issue:
35(6), P. 862 - 876
Published: March 21, 2024
Clustered
federated
learning
(CFL)
is
a
promising
solution
to
address
the
non-IID
problem
in
spatial
domain
for
(FL).
However,
existing
CFL
solutions
overlook
issue
temporal
and
lack
consideration
of
time
efficiency.
In
this
work,
we
propose
novel
approach,
called
ClusterFLADS
,
which
takes
advantage
false
predictions
inappropriate
global
models,
together
with
knowledge
temperature
scaling
catastrophic
forgetting
reveal
distributional
similarities
between
training
data
(of
different
clusters)
test
data.
Additionally,
design
an
efficient
feature
extraction
scheme
by
exploiting
role
each
layer
neural
network's
process.
By
strategically
selecting
model
parameters
using
PCA
dimensionality
reduction,
effectively
improves
clustering
speed.
We
evaluate
real-world
IoT
trace
various
scenarios.
Our
results
show
that
accurately
efficiently
clusters
clients,
achieving
$100\%$
true
positive
rate
low
positives
across
distributions
both
domains.
Sensors,
Journal Year:
2024,
Volume and Issue:
24(8), P. 2455 - 2455
Published: April 11, 2024
The
Internet
of
Things
(IoT)
is
what
we
have
as
a
great
breakthrough
in
the
5G
network.
Although
network
can
support
several
Everything
(IoE)
services,
6G
to
fully
that.
This
paper
survey
research
presenting
and
IoT
technology
challenges
coming,
with
being
new
alternative
coming
solve
these
issues
limitations
are
facing
5G.
A
reference
Control
Plane
User
Separation
(CUPS)
made
IPv4
IPv6,
addressing
which
foundation
slicing
for
core
In
comparison
other
related
papers,
provide
in-depth
information
on
how
going
affect
our
lives
this
handled
IoE
Finally,
full
network,
its
compared
IEEE Access,
Journal Year:
2022,
Volume and Issue:
10, P. 28975 - 28990
Published: Jan. 1, 2022
6G
technology
is
now
attracting
many
scientific
researchers
due
to
its
prominent
features
including
high
mobility,
data
rate,
operating
frequency,
and
ultra-low
end-to-end
delay.
Compared
the
5G,
has
certain
advantages
that
can
transform
smart
healthcare
a
new
advanced
intelligent
system,
where
multiple
issues
concerns
of
services
(e.g.,
slow
ambulance
response)
are
effectively
resolved.
In
such
communication
patients
service
providers
communicate
with
each
other
via
an
open
internet
channel.
Since
important
very
sensitive,
security
privacy
in
network
become
prominent.
this
paper,
we
introduce
6G-aided
environment.
Our
work
also
proposes
solution
called
Centerless
User-Controlled
Single
Sign-On
(CL-UCSSO)
for
achieving
convenient
cost-saving
multi-server
system
constructed.
A
three-factor
mechanism
(combining
card,
password
biometrics)
time
bound
property
integrated
design
protocol
fast
authentication
allows
efficiently
establish
secure
communications.
Security
proof
proposed
provided
using
well-known
verification
tools
RoR
model,
AVISPA
simulation
BAN
logic.
Results
performance
comparisons
on
various
aspects
show
our
provides
more
functionalities
incurs
less
cost
compared
related
works.
IEEE Transactions on Intelligent Transportation Systems,
Journal Year:
2023,
Volume and Issue:
24(4), P. 4158 - 4172
Published: Jan. 9, 2023
Falsification
detection
is
a
critical
advance
in
ensuring
that
real-time
information
about
vehicles
and
their
movement
states
certified
on
the
Internet
of
Vehicles
(IoV).
Thus,
detecting
nodes
are
propagating
inaccurate
requirement
for
successful
deployment
IoV
services
although
only
few
research
studies
have
been
carried
out
Basic
Safety
Message
(BSM)
falsification.
As
such,
this
paper
proposes
Randomized
Search
Optimization
Ensemble-based
Detection
Scheme
(RSO-FDS).
The
RSO
technique
was
used
to
construct
proposed
Random
Forest
(RF)
model.
evaluation
performed
three
different
datasets
developed
evaluate
falsification
IoV.
In
addition,
six
most
popular
supervised
learning
(SL)
algorithms
were
investigated
capability
RSO-FDS,
which
had
best
performance
across
all
datasets.
metrics
considered
computational
efficiency
terms
prediction
time,
validation
accuracy
overall
attack
classification,
precision,
recall,
F1
scores.
For
validation,
RSO-FDS
further
compared
with
results
from
recent
works.
Furthermore,
irrelevance
data
balancing
illustrated
real-life
scenarios.
result
shows
model
outperformed
state-of-the-art
implemented
work
related