Overlay Network Qos Routing

With the rapid development of Internet, a lot of new applications which have strict demands on quality of service have occurred in Internet, such as VoIP, network video conferences, network audio/video broadcast e.t.c. While the network layer of Internet only provides best-effort transport. It has poor error sufferance, can't provide satisifying QoS for these applications. Moreover, with more and more device with poor computating capacity (such as Mobile phone, PDA) are connected into Internet, people hope that most computation done in the network, the current networks can't provide the service. Overlay network is introduced to solve the above problems. It is a virtual network overlaid upon the existing networks, constructed with a lot of overlay nodes distributed in different ASes in Internet and the virtual links between each of them. It can easily be employed and provide much better quality of service with little change to the current network architecture.QoS routing on which this thesis focus is one of the key items of overlay networks research. In this thesis, the three stragegies of QoS routing is compared with each other, the advantages and disadvantages of each stragegy are discussed. Some typical QoS routing algorithms will be introduced briefly. Then the differences between IP QoS routing and QoS routing in overlay network are outlined. After an in-depth study of the exsisting overlay QoS routing solutions, we propose a generic overlay network architecture (GON), based on which a new solution for QoS routing is introduced, including topology construction, QoS routing Algorithm and maintenance of routing. Hierarchical routing strategy is adoptted in this thesis. A new QoS routing algorithm, named Resource-Balanced Least Cost Path (RBLCP) algorithm, is proposed for GON. With the new algorithm we can find a fleasible routing path as well as balancing the resource in the overlay network. Finally, the results of simulation show that RBLCP achives better performance in the feature of QoS success ratio than the exsisting corresponsive algorithms. It can also balance resource well, especially for the resource that is short in the network.