Optimization Of The Local Static Routing Algorithm On PDNS

With the explosive growth, people often meet a range of network issues, such as network congestion and low service quality. Therefore, it is necessary to analyze the large-scale network topology. As the primary tool of researching computer network, network simulation is widely used by the small cycle and operable feature. PDNS(Parallel and Distributed Network Simulator), as the popular distributed/parallel network simulation software, has greater advantages on simulating large-scale network. So the local static routing strategy which has great impact on the performance of PDNS is taken as the research object.The main research of local static routing strategy is routing computing on single simulator for static topology. By analyzing the typical local static routing strategies, the key technology is summarized on balancing the time complexity and space complexity from the current research of seven routing strategies. Routing memory requirement is still the bottleneck of large-scale simulation. This paper proposes the local static routing strategy bases on the zones by analyzing the network topology and implementing of routing on PDNS.This paper chooses a way to encode the nodes in the topology by using the tree shrinking, subnet partition and node coloring algorithms, and it can meet the space constraints according to the different features of the node vectors for the minimal visiting. By customized storage and searching schemes for different zones, this paper uses different routing method and least common ancestor algorithm to implement the routing computing and finding in order to balance the time and space complexity.Using the above strategies, experimental results show that the zone partition routing strategy can reduce the storage space for network simulation by 70% and equal to MTree_Nix routing on the simulation time which is more suitable for large-scale network simulation.