Research On Name Routing And Analysis Modeling For Source Management Routing Algorithm

Network routing has been the key issue in the network, today's computer networks are very large, high-speed, mass loaded with a variety of multimedia information, Therefore, the network routing is facing a new challenge. The goal of routing algorithm is to find the optimal path to transmit information; improve service quality. At present, the Internet routing algorithm is mainly based on the Dijkstra algorithm proposed in 1959 and Bellman-Ford algorithm proposed in 1962. These routing algorithms assume that the network topology is based on the definition of a metric, the shortest route between two nodes is calculated based on this metric, with maintaining a routing table entry for each sub-network. They are characterized by the global maintenance, precise routing, shortest path routing. As the networks are expanding and sharp increasing in the number of sub-network case, the current global optimal routing algorithms face serious challenges. Internet development trends lead to re-examine the existing routing algorithms and develop new routing algorithms.This paper summarizes the idea of name routing based on research on the VRR (Virtual Ring Routing) and ROFL (Routing on Flat Labels). Name routing is based on routing rules which is topology independent. It has non-precision routing, each node has only a finite number of points to reach the other nodes, the information is forwarded to the node who records the "recent" pointing to the node according to the routing rules; non-global route maintenance, each node only need to maintain the reachability of their own designated nodes.In this paper, we apply VRR(Virtual Ring Routing) and ROFL(Routing on Flat Labels) to support for BGP routing, and call this method as source management routing scheme. The following was found:the naming strategy impacts the routing performance for the source management routing scheme, although just naming randomly in VRR and ROFL. The naming strategy depending on topology can not improve the routing performance; the core ASes in the source management routing scheme play an important role in routing.We discuss the feasibility of source management routing scheme based on the actual the network topology with AS business relationships. The following was found: for directly employing source management routing scheme on the actual the network topology with AS business relationships, there are some routes that are not always meet the BGP path routing strategy; so we propose the hierarchical resource management routing scheme to establish and maintain multiple virtual neighbor paths in order to provide the complete routing information based on business relationships, to ensure the validity of the full path.We propose a theoretical model of the source management routing algorithm to describe the routing mechanism and the impacts on the routing performance. The routing algorithm is simple, still there is no theoretical model to describe how to choose the route and calculate the path length between a pair of nodes, especially in theory to discuss the performance of source management routing algorithm. The hop in virtual ring is this unique idea to describe the path of this routing algorithm. Using this model, how many hops in virtual ring to reach a destination can be calculated according to numerical distance between identifiers of two nodes. The probability of all possible paths in the virtual ring is calculated to get the average number of hops in the ring, and ultimately to estimate the actual physical path length. We analysize the factors that affect the routing performance, the number of nodes in the network, the virtual neighbor path length, the source node, destination node name and the network topology, all of them affect the source management routing path length distribution. Through a comprehensive in-depth analysis of source management routing scheme, we fully understand its routing efficiency and scalability, in order to adapt to the future core network and lay a solid theoretical basis.In order to solve the problem that the routing tables of the core nodes in source management routing algorithm are so big, we propose a heuristic address assignment algorithm based on probability to reduce the routing table size. Our algorithm can guarantee the high probability that each node uses the same next hop to the nodes assigned with consecutive addresses, and then these consecutive addresses can be compressed to compact the routing tables. Simulation results show that the probability-based heuristic algorithm can compress routing table greatly.