Constraint Driven - Multi Objective Cross-Layer Routing Protocols for VANETs

In recent years, Vehicular Ad-hoc Networks (VANETs) have become an important
and critical part of Intelligent Transport Systems (ITS). VANETs have introduced
a number of new services and potential applications for emergency, comfort and
commercial use for the transport community. Due to their infrastructure-less
and dynamic nature, VANETs have a number of design challenges. In particular,
designing a reliable, robust and efficient routing mechanism for VANETs is a
challenging task.

Conventional routing approaches for ad-hoc wireless networks, where each
layer of the model is optimised individually, do not cater for dynamic system
conditions encountered in VANETs. Primarily, in conventional routing schemes
each layer optimises itself without considering constraints from other layers. Such models work well with legacy services. However, with the invention of new applications and services, efficiency of these models lags behind the acceptable Quality of Service (QoS) standards.

In order to meet such QoS standards, this thesis proposes a novel cross-layer
routing paradigm for VANETs. In this proposed paradigm, considering the dynamic system conditions for VANETs, routing optimisation takes place across
multiple layers simultaneously. The cross-layer routing paradigm combines the
parameters from physical, network and medium access layers to make informed
routing decisions. The thesis explores integration of these parameters to cater
for application specific objectives for the multi-hop VANET environment. One of
the critical contributors in cross-layer design is VANET’s wireless channel. Due
to the dynamic and unpredictable nature of the wireless channel, its existing
models do not provide robust and efficient routing schemes. The thesis explores
the limitations of existing wireless channel models being used and proposes an
alternative model that captures more realistic environmental conditions.

Based on the proposed channel model, the thesis then formulates the proposed
cross-layer routing paradigm where routing decisions are based on varying channel conditions. Performance of the proposed routing architecture is compared with traditional routing schemes.

Extending the proposed routing framework, a data rate specific routing scheme
is introduced in the thesis. Performance of this scheme under a multi-hop VANET environment is examined. An analytical formation of the key performance metrics of this scheme is established. Furthermore, to cater for the application specific objectives, different variants of the proposed scheme are introduced in the thesis.

In order to formulate a more robust and reliable cross-layer routing mechanism,
the final stage of the thesis presents a joint rate and queue based routing solution. This model presents a constraint driven, multi objective routing paradigm where a number of possible influencing factors are integrated in decision making. Extensive experimental results demonstrate that such a cross-layer architecture performs better compared to existing schemes especially to cater for application specific requirements in a multi-hop VANET environment.