Information Technology Reference
InDepth Information
Neighbors are the nodes that are defined by the UDG model. The metric used to
measure routing optimality in this work is “Hop Count.” Hop count can be defined
as the number of time a data packet was forwarded on the created route from source
to destination node, ignoring potential retransmissions or acknowledgments [
15
].
An important definition is that of “progress” made while transferring the packet
from the one node to another. Let the source node (
S
) be the current node holding
the message, “
D
” the destination node, “A” the considered forwarding neighbor.
The progress made by forwarding data from
S
→
A
is
progress
=

rr rr
  
DS DA
−−
.
In other words, the UDG model will choose the next node among its neighboring
(onehop) nodes that maximizes this progress metric toward its destination [
15
].
7.6.3
Localized Probabilistic Progress Algorithm
This localized algorithm has been proposed in [
15
] where the probability of packet
reception is used to influence its neighbor selection. The
ppr
is calculated as
outlined earlier. The probabilistic localized algorithm is a simple extension of the
localized greedy algorithm where the neighbor selection criterion metric now
becomes
ppr
´
.
Nodes can record the
ppr
through successive measurements of uncorrupted
packets from neighbors and either assume that reciprocity holds, or inform their
neighbors of the measured
ppr
for each of their set of transmissions over a
predefined time window. Alternatively, a node can estimate rather than measure
ppr
from the received signal strength indicator (RSSI) of its radio receiver and employ
the intermediate signaltonoise and biterror probability models introduced in §7.3,
and once again either assume reciprocity or broadcast this information to each of
its respective neighbors.
In the simulations that follow we have assumed that the latter case (RSSI +
exchange of information) has taken place without any errors, which is an idealized
version of what is practically achievable.
progress
7.7
Numerical Simulation Setup
In this section we present briefly how we created a simulation environment for our
study. For the simulation we use a 2dimensional square simulation area and
uniformly and randomly distributed
n
nodes in this. The size of the simulation area
as well as the number of nodes are varied as described in the next section. The area
for the simulation and the number of nodes has been changed to cover a wide range
of neighboring node densities. In this study we assumed that all nodes are identical
Search WWH ::
Custom Search