In dynamic and wide networks, such as Metropolitan Mesh Network (MMN), routing becomes very complex because a packet can be blocked before it reaches its destination. In addition, users can also log in or log out from network topology. Therefore, a good routing algorithm, which is able to reduce time in network update process or when there is an error in the network, are required. Routing problems can be represented as the shortest path problem to facilitate completion. In this paper, a routing algorithm optimization using Adaptive Mutation Genetic Algorithm (AMGA) on MMN is presented by determining a probability of 0.000005782 at the beginning, with crossover probability of 0.000847, to reduce or avoid premature convergence.
Read full abstract