In WMNs, it is observed that the best routing technique is the one done
hop-by-hop as this technique adapts to the link status changes. Wireless links are
subject to changes in their bandwidth, interference level, etc. The effects of such
changes can be mitigated if the route establishment is done on the fly instead of
having it deterministic. So, self-routing helps in tackling the short-term path
quality variations problem of WMNs.
In VANETs, the most popular routing protocols are the self-routing ones as
they are the best handlers of the intermittent connectivity and dynamic topology.
Derived from the summarizing figures presented in the previous chapter,
Fig. 4.1 illustrates the ideal model and its ideal functionality per each component.
Self-routing requires that route discovery is handled in a reactive way, route
selection should be intermediate-based, and the route representation and data
forwarding is done based on route guidance.
For the selection metric, each network paradigm should have its own metric
based on the goals and needs of each paradigm. MANETs can have the mobility
level as its selection metric. Choosing the most stable neighbor-the slowest one-to
be the next forwarding node reduces the probability of route failures due to links
changes. For WSNs, the selection metric can be the residual energy of the
potential forwarders. Choosing the neighbor with the highest residual energy
reduces the chance of the node's battery depletion. In WMNs, there are different
selection metrics that can be utilized and are suitable for the intermediate-based
selection. These metrics are discussed in Table 3.1 and we can refer to them here
as QoS level. Finally, the most popular self-routing based VANET protocols are
based on greedy selection; in other words, the selection metric for VANETs should
be the distance to the destination.
(Selection Metric: mobility level [MANETs], residual energy [WSNs], QoS level [WMNs], distance to
the destination [VANETs])
Route Representation and Data Forwarding
Ideal routing model