IEEE Access,
Год журнала:
2023,
Номер
11, С. 134074 - 134086
Опубликована: Янв. 1, 2023
Due
to
the
widespread
use
of
mobile
intelligent
terminal
devices,
Mobile
Crowd
Sensing
(MCS)
applications
have
gained
significant
research
attention.
However,
ensuring
users
privacy
remains
a
critical
challenge,
as
it
can
hinder
users'
willingness
participate
actively
in
tasks.
To
address
limitations
existing
differential
protection
methods,
this
paper
proposes
novel
approach
based
on
Artificial
Immune
Computing
(AICppm).
Specifically,
private
information
is
concealed
within
masking
carrier,
and
data
scrambling
avoided.
The
proposed
method
involves
two
main
steps:
first,
carrier
preprocessing
high-pass
filter
bank
designed
identify
candidate
positions
for
perturbation.
Then,
steganography
algorithm
multi-objective
optimization
used,
transforming
perturbation
position
into
an
antibody
using
artificial
immune
algorithm.
By
iteratively
searching
antibodies
with
higher
fitness,
optimal
offspring
population
identified
improved
Strength
Pareto
Evolution
Algorithm
(SPEA2).
experimental
results
demonstrate
that
withstand
attacks
malicious
steganalysis
tools,
preserving
integrity
sensing
enabling
real-time
processing
information.
IEEE Transactions on Consumer Electronics,
Год журнала:
2024,
Номер
70(1), С. 2303 - 2310
Опубликована: Фев. 1, 2024
This
paper
proposes
a
federated
learning-based
decision
making
framework
for
sustainable
irrigation
using
IoT
and
dew-edge-cloud
paradigm.
The
learning
is
used
to
prevent
the
sharing
of
user
identities
raw
data
privacy
protection.
Further,
gradient
encryption
leakage
information.
Long
short-term
memory
(LSTM)
network
deep
neural
(DNN)
are
analysis
in
local
global
models.
Edge
computing
reduce
energy
consumption
latency.
cache-based
dew
provide
temporary
holding
when
connectivity
not
available.
results
present
that
proposed
achieves
~99%
prediction
accuracy
at
~50%
lower
latency
than
conventional
edge-cloud
framework.
Sensors,
Год журнала:
2025,
Номер
25(5), С. 1601 - 1601
Опубликована: Март 5, 2025
Using
mobile
crowd
sourcing/sensing
(MCS)
noise
monitoring
can
lead
to
false
sound
level
reporting.
The
methods
used
for
recruiting
phones
in
an
area
of
interest
vary
from
selecting
full
populations
randomly
a
single
phone.
Other
apply
clustering
algorithm
based
on
spatial
or
parameters
recruit
MCS
platforms.
However,
statistical
t
tests
have
revealed
dissimilarities
between
these
selection
methods.
In
this
paper,
we
assign
(1)
acoustic
characteristics
and
(2)
outlier
affecting
the
level.
We
propose
two
phases
approach
starts
by
applying
form
focused
clusters
removing
outliers.
Then,
is
applied
eliminate
This
creates
subsets
that
are
calculate
conducted
real-world
experiment
with
25
performed
test
evaluation
methodologies.
values
indicated
dissimilarities.
compared
our
proposed
method
terms
properly
detecting
eliminating
Our
offers
4%
12%
higher
performance
than
method.
ACM Transactions on Sensor Networks,
Год журнала:
2025,
Номер
unknown
Опубликована: Март 11, 2025
In
pursuit
of
an
immersive
virtual
experience
within
the
Cyber-Physical
Immersive
Networking
Systems
(CPINS),
construction
scenarios
often
requires
a
considerable
amount
real-world
data.
Mobile
Crowd
Sensing
(MCS)
represents
one
effective
methods
during
data
collection
Metaverse.
However,
privacy-preserving
and
quality-aware
are
two
critical
contradictory
issues
in
MCS,
because
hiding
as
much
personal
information
about
users
possible,
while
learning
possible
to
recruit
high-quality
for
collection.
To
this
end,
we
propose
Privacy-Preserving
Reputation
Calibration
based
Quality-aware
Data
Collection
(PPRC-QDC)
scheme.
PPRC-QDC
scheme,
two-tier
truth
discovery
is
proposed
acquire
More
importantly,
method
recognize
users’
reputations
by
comparing
weights
with
trusted
rather
than
effectively
identify
honest
users.
Finally,
theoretical
analysis
confirms
our
has
stronger
privacy
preservation
robustness
capability.
Extensive
experiments
conducted
on
datasets
demonstrate
that
under
preservation,
scheme
recognizes
accuracy
91.5%,
improves
quality
11.0%-12.1%.
Proceedings of the ACM on Human-Computer Interaction,
Год журнала:
2025,
Номер
9(2), С. 1 - 30
Опубликована: Май 2, 2025
Mental
health
is
a
growing
concern,
especially
among
young
adults,
but
gathering
data
from
this
demographic
presents
distinct
challenges.
Crowdsensing
research
approach
that
has
become
increasingly
popular
due
to
its
ability
collect
many
individuals
continuously
and
at
scale.
However,
it
equally
important
ensure
the
collected
of
high
quality,
as
depends
on
factors.
In
paper,
we
discuss
quality
issues
encountered
during
our
crowdsensing
study
conducted
October
2022
August
2023,
which
aimed
collecting
college
students'
emotions
mental
wellness.
We
present
findings
related
participant
recruitment,
device
usability,
quantity,
compliance,
consistency,
privacy
concerns,
incentive
mechanisms.
strategies
address
these
challenges
plans
for
future
improvements.
Our
results
discussion
highlight
effectiveness
in
collection
demographic.
Additionally,
identified
positive
negative
emotional
drivers
potential
stressors
affecting
group's
The
insights
work
can
aid
design
applications
studies.
IEEE Access,
Год журнала:
2023,
Номер
11, С. 136150 - 136165
Опубликована: Янв. 1, 2023
In
today's
world,
the
importance
of
Green
Internet
Things
(GIoT)
in
transformed
sustainable
smart
cities
cannot
be
overstated.
For
a
variety
applications,
GIoT
may
make
use
advanced
machine
learning
(ML)
methodologies.
However,
owing
to
high
processing
costs
and
privacy
issues,
centralized
ML-based
models
are
not
feasible
option
for
large
data
kept
at
single
cloud
server
created
by
multiple
devices.
such
circumstances,
edge-based
computing
used
increase
networks
bringing
them
closer
users
decentralizing
without
requiring
central
authority
circumstances.
Nonetheless,
enormous
amounts
stored
distribution
mechanism,
managing
application
purposes
remains
difficulty.
Hence,
federated
(FL)
is
one
most
promising
solutions
end
devices
through
edge
sharing
private
with
server.
Therefore,
paper
proposes
learning-enabled
system,
which
seeks
improve
communication
strategy
while
lowering
liability
terms
energy
management
security
transmission.
The
proposed
model
uses
FL
produce
feature
values
routing,
could
aid
sensor
training
identifying
best
routes
servers.
Furthermore,
combining
FL-enabled
techniques
simplifies
also
allowing
more
efficient
system.
experimental
results
show
an
improved
performance
against
existing
network
overhead,
route
interruption,
consumption,
end-to-end
delay,
interruption.
IEEE Access,
Год журнала:
2023,
Номер
11, С. 140325 - 140347
Опубликована: Янв. 1, 2023
Due
to
the
growing
capabilities
of
mobile
phones
and
devices,
crowd
sensing
(MCS)
is
rapidly
gaining
popularity
among
researchers
in
different
fields,
given
its
ability
collect
data
at
scale
low
cost.
MCS
particularly
important
healthcare
domain
since
it
provides
opportunities
health,
wellness,
Quality
Life
information
from
a
large
diverse
population.
For
example,
can
be
used
detect
early
signs
emerging
health
conditions,
track
spread
infectious
diseases,
assess
effectiveness
interventions,
without
need
for
frequent
clinical
visits.
Consequently,
also
reduce
costs
help
overcome
barriers
access.
This
article
takes
closer
look
systems
that
have
been
research
medical
domains.
We
provide
thorough
analysis
selected
based
on
their
health-related
objectives,
such
as
monitoring
physical
activity,
detecting
preventing
disorders,
providing
treatment.
adopt
three-layered
architecture
structure
health-centric
frameworks,
consisting
application,
data,
layers.
In
application
layer,
we
analyze
participant
recruitment,
incentive
mechanisms,
task
allocation
strategies.
types
collected
how
they
are
stored
processed
future
use.
The
layer
specifies
methods
explains
fundamental
requirements
lower
level.
Additionally,
explore
significant
challenges
faced
by
existing
domains
offer
promising
avenues
research,
which
user
privacy,
resource
utilization,
quality,
compliance.
work
insights
into
some
practical
applications
MCS,
highlights
solutions,
addressed,
all
catalyze
development.
The
collection
and
transportation
of
samples
are
crucial
steps
in
stopping
the
initial
spread
infectious
diseases.
This
process
demands
high
levels
safety
timeliness.
rapid
advancement
technologies
such
as
Internet
Things
(IoT)
blockchain
offers
a
viable
solution
to
this
challenge.
To
end,
we
propose
Blockchain-enabled
Infection
Sample
Collection
system
(BISC)
consisting
two-echelon
drone-assisted
mechanism.
utilizes
collector
drones
gather
from
user
points
transport
them
designated
transit
points,
while
deliverer
convey
packaged
testing
centers.
We
formulate
described
problem
Two-Echelon
Heterogeneous
Drone
Routing
Problem
with
Transit
point
Synchronization
(2E-HDRP-TS).
obtain
near-optimal
solutions
2E-HDRP-TS,
introduce
multi-objective
Adaptive
Large
Neighborhood
Search
algorithm
for
(ALNS-RD).
algorithm’s
functions
designed
minimize
total
time
infection
exposure
index.
In
addition
traditional
search
operators,
ALNS-RD
incorporates
two
new
operators
based
on
flight
distance
index
enhance
efficiency
safety.
Through
comparison
benchmark
algorithms
NSGA-II
MOLNS,
effectiveness
proposed
validated,
demonstrating
its
superior
performance
across
all
five
instances
diverse
complexity
levels.
