Cohen, Joshua Goldberg, and Jaudelice C. de Oliveira
than cubic, which is the realistic scenario. Therefore, our encoding technique
is much simpler and suitable for deployment in actual sensor networks.
Table 1 summarizes the di?®erences between our proposed algorithm (TER)
and the algorithm described in [14, 15, 16]. The following conclusions can be
drawn from the table:
Table 1. Comparison between TBF and TE Routing (TER) mechanisms.
Metric TBF TER Comparison-TBF Comparison-TER
Path Definition Source-Defined.
Continuous
parametric
function.
Sink-Defined.
Arbitrary function
defined by the
expert.
Dynamic Curve
Definition.
Lack of Curve
Complexity.
Packet Overhead.
Initial point
distribution
overhead.
Expert defined
curve.
Fixed curve
granularity.
Curve Breakdown Every hop breaks
the curve down
into discrete
points.
Calculation done
for each
neighborhood
node.
Sink breaks the
entire curve down
into discrete
points.
The sink
distributes all of
the points to the
nodes.
Static breakdown
at the sink.
Dynamic
breakdown at
nodes. Energy can
be an issue.
Closeness to Curve Curves which form
loops will skip
sections.
Global discrete
breakdown
enumerates all the
points.
The traveresed
path will pass over
every discrete
point when
possible.
Lacks support for a
specific curve type.
Supports accurate
following of the
specified curve.
Success Rate Best possible success
rate given onehop
neighbor information
available.
Pages:
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608