Supporting Internet Access and Quality of Service in Distributed Wireless Ad Hoc Networks

In this era of wireless hysteria, with continuous technological advances in wireless communication and new wireless technologies becoming standardized at a fast rate, we can expect an increased interest for wireless networks, such as ad hoc and mesh networks. These networks operate in a distributed manner, independent of any centralized device. In order to realize the practical benefits of ad hoc networks, two challenges (among others) need to be considered: distributed QoS guarantees and multi-hop Internet access. In this thesis we present conceivable solutions to both of these problems. An autonomous, stand-alone ad hoc network is useful in many cases, such as search and rescue operations and meetings where participants wish to quickly share information. However, an ad hoc network connected to the Internet is even more desirable. This is because Internet plays an important role in the daily life of many people by offering a broad range of services. In this thesis we present AODV+, which is our solution to achieve this network interconnection between a wireless ad hoc network and the wired Internet. Providing QoS in distributed wireless networks is another challenging, but yet important, task mainly because there is no central device controlling the medium access. In this thesis we propose EDCA with Resource Reservation (EDCA/RR), which is a fully distributed MAC scheme that provides QoS guarantees by allowing applications with strict QoS requirements to reserve transmission time for contention-free medium access. Our scheme is compatible with existing standards and provides both parameterized and prioritized QoS. In addition, we present the Distributed Deterministic Channel Access (DDCA) scheme, which is a multi-hop extension of EDCA/RR and can be used in wireless mesh networks. Finally, we have complemented our simulation studies with real-world ad hoc and mesh network experiments. With the experience from these experiments, we obtained a clear insight into the limitations of wireless channels. We could conclude that a wise design of the network architecture that limits the number of consecutive wireless hops may result in a wireless mesh network that is able to satisfy users’ needs. Moreover, by using QoS mechanisms like EDCA/RR or DDCA we are able to provide different priorities to traffic flows and reserve resources for the most time-critical applications.