Image Processing Reference
In-Depth Information
other protocols. An interesting approach is proposed in [Lin], where the energy
delay metric is
suggested. As energy consumption strongly depends on the amount of data exchanged, in some cases
the overhead due to route setup and management is also used to assess the performance of a WSN.
Another quite popular metric is the time to get the network partitioned. his is especially impor-
tant in algorithms that aim to maintain network connectivity as long as possible. To assess their
performance, optimization-based routing algorithms may use the same metric optimized by the cost
function, e.g., the energy consumed along the chosen paths. However, the most important objec-
tive of a routing protocol is usually to maximize the network lifetime rather than minimize the
energy of single paths. So the most widely used metrics refer to the network lifetime, which can
be expressed in several different ways. For example, in WSNs used for intrusion or fire detection,
where network quality decreases considerably as soon as one node dies, the first node dies (met-
ric, which gives an estimated value for this event for a specific network configuration, is appropriate
[HAN]. In scenarios where the loss of a single or a few nodes does not automatically compro-
mise network operation, on the other hand, the half nodes alive metric, which gives an estimated
value for the half-life of a WSN, can be used. Finally, the last node dies metric gives an estimated
value for the overall lifetime of a WSN. These metrics can be expressed in terms of either time
or rounds, where such a parameter is defined (e.g., in Low Energy-Adaptive Cluster Hierarchy,
LEACH [Hei]).
Other metrics may depend on the protocol category, e.g., quality of service (QoS) metrics for
QoS-enabled protocols, such as deadline hit and miss ratio for real-time routing.
∗
7.1.4 Role of Topology Management Protocols
Energy consumption in WSNs is typically dominated by the node communication subsystems, and
can only be significantly reduced by transitioning the embedded radios to a sleep state, at the expense
of changes in the network topology. The role of topology management protocols is to select which
nodes can turn off their radios without compromising the network capacity.
To this end they coordinate the sleep transitions of all the nodes, while ensuring adequate network
connectivity, in such a way that data can be efficiently forwarded to the data sink.
The following parts of this chapter are organized as follows. Section . presents an overview and
classification of energy-efficient routing protocols for WSNs. Sections . through . describe and
discuss in detail notable representatives of power-efficient routing protocols. For ease of presenta-
tion, each protocol is presented in the context of the category it belongs to. In Section . topology
management schemes for WSNs are addressed. Finally Section . outlines open issues in energy-
efficient routing protocols for WSNs.
7.2 Overview of Energy-Saving Routing Protocols for WSNs
Energy-saving routing protocols for WSNs can be classified into five main categories, i.e.,
optimization-based, data-centric, cluster-based, location-based, and QoS-enabled. Such categories
are not necessarily disjoint, and some examples of routing algorithms matching multiple categories
canbefound.
7.2.1 Optimization-Based Routing Protocols
A broad spectrum of routing algorithms for WSNs aiming at reducing the energy consumption of
sensor nodes is present in the literature. Some of them take energy into account explicitly when
routing sensor data, and for most of them the main goal is the optimization of some metric. For
this reason, we will henceforward refer to them as optimization-based energy-aware routing (EAR)
