Study Of Distributed QoS Routing Algorithm

Current internet can only provide "best-effort" service, which means it will try its best to forward user's data packets, but can provide no guarantees regarding bandwidth and delay. While this kind of service is suitable for some traditional applications, it's intolerable for newly emerged real-time, multimedia applications. How to provide high quality of service in the next generation internet is a key problem in computer network research.There are many aspects which networks provide the guarantee of the quality of service. However the one of the key technology for providing the guarantee of the quality of service is the QoS routing. Researchers and experts propose many creative and practicable QoS routing algorithms, which can be divided into two classes: source routing algorithms and distributed routing algorithms. Though having some defects, these algorithms contribute a lot to network researches.This thesis proposes to design and realize distributed QoS routing algorithms: Distributed QoS Routing Algorithm with Multi-path Probing (MPDQRA) and Distributed Parallel QoS Routing Algorithm with Multi-path Probing (MPDPQRA). In the MPDQRA, every node only need to maintain local state information of its adjacent nodes, the metric of which is success ratio of node is used to decrease the extension of flooding, while Multi-Path can increase the probability of finding the feasible paths. Based on improved MPDQRA, MPDPQRA probes and reserves resources in one of several parallel paths according to the status of network and the connection QoS requirements. MPDPQRA can reduce resource fragments and admit more services into a network with heavy load.The thesis creates network simulate scenes and go along simulate experiment using these two QoS routing algorithms. By comparison with traditional shortest path first routing algorithm, it can be seen that these two algorithms can balance network load and promote network resource utilization. Under the circumstance of the network light load, their communication overhead is smaller. Under the circumstance of the network overload, MPDPQRA can combine with several paths to admit new connection, so it improves the network resource utilization ratio and throughput. In conclusion, applying the proposed algorithms to connection-oriented multimedia data network will improve the capability of transmitting multimedia information.