Design and software implementation of routing in networks

number: 
784
إنجليزية
Degree: 
Author: 
Mohammed Salah Saleh Zeki
Supervisor: 
Dr. Siddeeq Y. Ameen
year: 
2002
Abstract:

One of the main problems in computer networks is to find the best route from source to destination. Many routing algorithms were proposed, and implemented in router to solve this problem. To implement efficient routing algorithm in computer network, Social Network Analysis (SNA) was used first to find the optimum network topology. Two design strategies for social network analysis were adopted in this work. The proposed strategy of increasing links was implemented by two different networks (directed and undirected), and the effect of nodes and links failure on the centrality measures and reliability has been applied again. The optimization process leads to provide robust topology and enough geodesies path for the network. Also, the optimized topology will be used to implement the routing simulator. Second, efficient routing algorithm (link state routing algorithm) has been adopted and simulated in this work. It was based on the concept of multithreading. These feature which provided by Java language can provide sharing between threads. The designed program can runmultiple threading at the same time and can easily f nd the final routing table for each router in different networks (directed/undirected), and include more than parameters. The improved networks using SNA are implemented and tested using link state algorithm. The maximum available shortest paths for each router in the network by using the proposed learning factor which objective the adoptively of link state, this can be done by initiate the learning for the previous rouier to stop use the path selected first and use the next shortest path. The effects of cost change and link failure has been simulated and iscussed. The simulator will chose the shortest path with minimum cost not hop. In addition, the scheduling and priority concept was implemented in this work. Finally, the problem of link capacity was implemented by controlling the capacity of each link in the network topology for different case studies, this case reduce and split the traffic of packets imposed on the shortest cost path.