Image Processing Reference
In-Depth Information
protocols. Example of metrics to be minimized are the energy consumed per message, the variance
in the power level of each node, the cost/packet ratio, or the maximum energy drain of any node.
Trying to minimize the energy consumed per message may lead to poor routing choices, as some
nodes could be unnecessarily overloaded and thus could quickly extinguish their batteries. A more
effective option, if all nodes are equally important for the WSN to operate correctly, is to try to balance
the battery power remaining in the nodes, as there is no point in having battery power remaining in
some nodes while the others have already run out of power.
Minimization of the cost/packet ratio involves labelling different links with different costs and then
choosing the best option so as to delay the occurrence of network partitioning as long as possible.
On the other hand, the idea of minimizing the maximum energy drain of any node derives from
the consideration that network operations start to be compromised when the first node exhausts its
battery, so it is advisable to minimize battery consumption in this node.
A number of optimization-based power-aware routing approaches try to maximize network life-
time. hey target network survivability, meaning that their goal is to maintain network connectivity
as long as possible. To achieve this goal, “optimal” routes that avoid nodes with low batteries and
try to balance the traffic load are chosen [Cha]. The use of optimization techniques to find the
minimum-cost path, where the cost parameter takes energy (alone or combined with other metrics)
into account, is proposed. However, the minimum cost path approach has a drawback in terms of
network lifetime in the long term. In fact, a protocol which, once it has found an optimal path, uses
only that path for routing, will eventually deplete the energy of the nodes along the path. As large
differences in the energy levels of the WSN nodes could lead to undesired effects such as network par-
titioning, suitable solutions have been developed. A notable example is the EAR protocol [Sha],
where network survivability is pursued by choosing not a single optimal route, but a set of good
routes, i.e., suboptimal paths which are selected in a probabilistic way. Section . will present and
discuss this approach.
7.2.2 Data-Centric Routing Protocols
Unlike the optimization-based routing algorithms described above, other routing protocols for WSNs
obtain low-power consumption for sensor nodes without explicitly dealing with energy considera-
tions when performing route selection, but implementing mechanisms which reduce energy wastage.
One of the main causes of energy wastage in WSNs is data redundancy, which derives from a com-
bination of a lack of global identifiers (as no IP-like addressing is possible in WSNs) and the random
deployment of sensors, which in many cases makes it difficult, if not unfeasible, to select a speci-
iedsetofsensorswithinagivenarea.Tosolvethisproblem,data-centricroutingapproacheswere
introduced. In these approaches, data is named using high-level descriptors, called meta-data, and
data negotiation between nodes is used to reduce redundancy. Another approach to reduce data
redundancy (and the consequent energy wastage) is by performing data aggregation at the relay-
ing nodes, which consists of combining data from different sources and eliminating duplicates, or
applying functions such as average, minimum, and maximum. Data aggregation also overcomes the
overlap problem, which arises when multiple sensors located in the same region send the same data
to the same neighbor node. Thanks to data aggregation significant energy savings can be achieved,
as computation at sensor nodes is less energy-consuming than communication. When performed
through signal processing techniques, data aggregation is referred to as data fusion. According to the
kind of routing protocol, data aggregation may be a task performed by special nodes or any node
in the network. Notable examples of data-centric routing protocols which perform data aggregation
for energy-saving purposes are sensor protocols for information via negotiation (SPIN) [Kul] and
Directed Diffusion [Int], which in turn inspired several other protocols. hey will be discussed in
Section ..
