Path Selection Optimization using Genetic Algorithm in IP Network

A novel approach of optimized routing which is based on Genetic Algorithm has been introduced. It finds out the optimum and the most suitable path which qualifies Objective Function criteria. Optimization Approaches have been applied to various real life issues in communication and networking. The path which is best fitted in the population is considered as the optimal path. It is obtained after qualifying the fitness function measuring criteria. The fitness function measures the best fitted path based on constraints; bandwidth, delay, link utilization and hop count. Population is composition of valid and invalid paths. In this research work we proposed a new genetic algorithm for network path optimization. Network path selection is based upon the fitness function measuring average packet delay for a network path. The population comprises of all chromosomes of variable length, so that the algorithm can perform efficiently in all scenarios. Rank-based selection is used for cross-over operation. Mutation operation is used for maintaining the population diversity. The results prove our assertion that our proposed algorithm finds the optimized shortest path between source nodes to destination node more efficiently than existing algorithms.