Abstract: Aimed at increasingly complex network scales, existing routing algorithms do not select the best path for packet forwarding, ant colony optimization routing algorithm (ACO-RA) is proposed. First, virtual network environment is designed on NS-2 platform, and the communication bandwidth and distance between each link are randomly initialized, and the link bandwidth is taken as the pheromone on the ant colony search path. Then according to ant colony optimization algorithm's path selection mechanism, it dynamically finds the best path for packet forwarding. Finally, experimental simulation is performed by otcl programming, the convergence time and the optimal path length of the network packet forwarding are used as performance indicators. The advantages and disadvantages of ACO-RA, DV, and SPF algorithms for network packet path selection are compared and analyzed, proved the advantages of the ACO-RA algorithm.

Key words: ant colony optimization algorithm, route algorithm, packet forwarding, distance vector algorithm, short path first algorithm