SEARCH
0-9 A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Prev | Current Page 597 | Next

Yingshu Li, My T. Thai, and Weili Wu

"Wireless Sensor Networks and Applications"


Sub-optimal
performance since
only a portion of
single hop
neighbors are
considered.
All available information
is used to
complete the message
path.
The search window
limits which of the
neighboring nodes
will be considered.
Computation
Complexity
P points per hop:
P function
evaluations.
M nodes in the
neighborhood.
P distance
calculations :
P ?¤M?¤(2MULT +
1ADD/SUB).
Define the search
window per hop:
16ADD/SUB +
17MULT.
N nodes in the
search window.
N distance
calculations :
N ?¤ (2MULT +
1ADD/SUB).
Depending on how
the next hop is
chosen, P can be
small.
N will be smaller
than P*M. The
search window is
only calculated
once and is comparable
to P function
evaluations.
??? TBF requires the source to define the path as a continuous function in
t, while in TER the path is defined, regardless of complexity, at the sink
and distributed as a set of discrete points to the nodes. TBF may allow
for a di?®erent path to be used for every packet. This comes at the cost of
overhead and lack of support for complex curves.
??? TBF requires that every node discretizes the curve within its neighborhood.
The number of calculations is dependent on dt and the complexity
of the curve function. In TER, the curve is discretized a priori. The curve
only needs to be discritized once in either case, but discretizing it at the
sink saves power and allows for more complex curves.
??? TBF requires floating point calculations to discretize the curve.


Pages:
585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609