| The fast technological progress and innovations of the past years have taken wireless communication from concept to reality. Indeed, wireless networks are now very well operating with tremendous advantages. There is no more need for expensive infrastructure to communicate efficiently. With ad hoc networks technology, mobile devices are free to move and communicate wirelessly with one another. Many important and sometimes vital applications are supported by this technology; disaster-and-relief operations, emergency rescue, law enforcement, and military battlefields are just few of them. However, one of the most challenging problems in ad hoc networks research is to ensure good quality voice and video multimedia services. These kinds of services need specific requirements to be satisfied in terms of bandwidth, end-to-end delay, jitter, etc., and this is very difficult to achieve in ad hoc networks, because of the topology dynamics and the scarcity of resources that characterize them.;The objective of this thesis is to investigate new approaches towards achieving good quality of service in wireless ad hoc networks. With the aim of obtaining more adaptive and flexible models, we propose a set of intelligent solutions based on neural networks and fuzzy logic techniques. These artificial intelligence tools have proved their benefits and efficiency in both academia and industry. The simulations we realized show that the proposed quality of service approaches promise to be efficient and reliable solutions for supporting both real-time services and non real-time applications. The obtained results illustrate that the intelligent-based techniques can raise successfully the challenge of quality of service satisfaction in wireless ad hoc networks. |
