Image Processing Reference
In-Depth Information
reaches the destination. Route maintenance is very simple, as it just sporadically performs localized
flooding in order to keep all the paths alive.
EAR requires a lower overhead than Directed Difusion. Simulation results show that the improve-
ment in power consumption is up to .%, while the network lifetime, here meaning the time when
the first node runs out of energy, increases up to %. However, these results assume that energy
is only consumed for packet transmission and reception, i.e., there is no energy consumption dur-
ing the idle state. Thus, also in EAR, energy efficiency is accomplished by reducing the overhead.
While the routing protocol does not explicitly decrease either the duty cycle or collisions, the authors
propose the design of a sensor node that follows the PicoRadio architecture [DaS], which features
two radio transceivers, one always active but with a very low bit rate and power consumption, and
the other with a low duty cycle and a higher data rate. With a similar node, CSMA/CA can be used
in the low-power signalling channel in order to avoid collisions during data transmission.
7.4.4 Maximum Lifetime Routing Protocol
In [Cha] the routing problem in a WSN is modeled as a linear optimization problem. Instead of
trying to minimize the energy of single transmissions or single paths, the objective is maximization
of the lifetime of the overall sensor network. his work extends previous work by the same authors
presented in [Cha] and [CT]. In these works it was found out that, in order to maximize network
lifetime, it is more useful to balance the energy consumption between the nodes in proportion to their
energy rather than minimize the power consumed by transmissions. he algorithm used to balance
energy along the whole path is the Maximum Residual Energy Path. This algorithm always selects
the path with the maximum residual energy to route packets across the network. Nodes can calculate
the maximum residual energy path in several different ways. In [CT] the use of a path length vector
is proposed, which works as follows: for each path from a source
i
and a destination
d
, a path length
L
p
is defined that is a vector of link costs
c
jk
,where(
j
,
k
)isalinkinthepath.hevalueof
c
jk
is the
reciprocaloftheresidualenergyatnode
j
after the routing has been performed, i.e.,
C
jk
=
(.)
E
j
−
e
jk
where
E
j
istheresidualenergyatnodej
e
jk
is the energy consumed to transmit over the link
(
j
,
k
)
Usingthisformulation,themaximumresidualenergypathcanbecalculatedastheshortestpath
with a slightly modified version of the Bellman-Ford algorithm [Bel] [For] which compares the
entries of the path length vector in lexicographical order. In [Cha] an alternative way to calculate
the maximum residual energy path is also provided, simply using the largest element of the path
length vector for comparison. As an alternative, the link cost is proposed, which reflects the number
of packets that can be delivered with the residual energy of the nodes, so the cost value is
e
ij
E
i
C
ij
=
.
(.)
The protocol has been compared with the minimum transmitted energy (MTE) algorithm [She] (in
whicheachnodesendsamessagetotheclosestnodeonthewaytothebasestation,BS)thatuses
e
ij
as
the transmission cost. Simulation results show that the Maximum Residual Energy algorithm is more
energy-efficient than MTE, as the network lifetime noticeably increases. Furthermore, the new metric
is shown to perform better than the one presented in [CT], but the use of the maximum cost instead
of the whole vector of path costs causes a slight performance degradation
.
So the best combination is
the new metric combined with the old way to calculate the maximum residual energy path. However,
