Research On Network Optimization Control Based On Network State Perception

With the rapid development of IP internet, the requirements of the internet on the network Quality of Service( Qo S) are also more and more complex, the type and the number of the internet is growing rapidly, the relationship between the various network business and the internet becomes more and more complex. It has a big significance to ensure the end-to-end performance of the network is most optimal and to achieve a reasonable distribution of the network.This article mainly made a study and improvement on the Chaos Ant Colony Optimization(CACO), and then the improved CACO is applied to the routing optimization of network to achieve the network optimization control. The main contents are as follows:(1) Firstly, the paper made an analysis of network optimization on the present situation, research background and research significance of both at domestic and international, studied on the optimizing indexes of network and the key technology of network routing, at the same time, in order to realize the network optimization control, the network state aware method and the selection of network state parameters were analyzed and discussed.(2) Secondly, this paper made an analysis on the Ant Colony Optimization(ACO) and the chaos theory, an improved CACO is given by combining chaos theory and ACO while a new formula of updating pheromones was also introduced at the same time. The improved algorithm avoid the defects which the ACO was easy to fall into precocity, stagnation, and long optimization time by using the characters of chaotic motion’s stochastic, ergodicity and sensitivity to initial conditions, the introduction of new formula of updating pheromones enhanced the resolution of high quality solutions and the simulation results proved the validity and reliability of the algorithm.(3) Finally, the paper made a study on network model and the network routing optimization problem, and aiming at the optimization of multi-objective problem, a network routing optimization method is given on the base of the improved Chaos Ant Colony Optimization algorithm.. This method realized the load balance and the reasonable allocation of network resources on the base of meeting requirements of bandwidth, delay, delay jitter, packet loss rate and cost, the simulation indicates that the algorithm is feasible.