Image Processing Reference
In-Depth Information
TABLE .
A Summary of Distributed Data Extraction Techniques for WSNs
Characteristics
Essential Conceptual Features
Platform
Name of the Project/
In-Network
Acquisitional Query
Cross-Layer
Data-Centric Query/
Mechanism Author
Processing
Processing
Optimization
Data Dissemination
Simulation
Mote-Class
PDA-Class
PC-Class
Directed diffusion []
Uses filters
×
×
Publish/subscribe applica-
tion programming inter-
face (API) based on named
data
✓
✓
Suppression of
duplicate messages
Gradients used to route data
from source to sink
Cougar []
Aggregation at cluster
leaders
×
×
×
✓
Packet merging
TinyDB []
Evaluation of
operators such as
MIN, MAX, COUNT,
AVERAGE, etc.
Centralized query
optimization based on
collected metadata
Communication
scheduling primarily
for upstream
data flow
Semantic routing
✓
✓
Priority given to cheapest
sensor when evaluating
multiple predicates
Nodes able to decide on
schedule themselves
using hop count
Designed for range queries
Multi-query optimization:
Grouping of multiple
identical events
Minimizes burden
on MAC
Maintenance algorithm
may be too costly if
measured attribute
changes too rapidly
Scheme unusable for
multiple concurrent
queries with different
epoch requirements
Bonfils et al. []
Decentralized query
operator placement
×
×
×
✓
(
continued
)
