| With the rapid development of Internet, Internet displays more and more multimedia applications, which leads to an increasing demand for the performance of the network. Traditional network that "best" is transmitted has been unable to meet such a high demand for data transfer performance. In this condition, quality of service (QoS) routing is proposed by industry and evolved into one of the Web’s core technologies. Internet applications are very complex, and there are many constraints, on this premise QoS routing problem becomes a NP-Complete problem, which is difficult to solve using traditional methods.Ant Colony algorithm,also known as Ant Colony algorithm, is made from natural biological Bionic optimization algorithm of swarm intelligence, with features such as self-organization, parallelism, and flexibility. Although compared to other smart algorithms, it seems to appear relatively late, but in solving complex optimization problems, especially discrete and multidimensional dynamic optimization problems, the advantage of the algorithm shows.This paper introduces the basic concepts of QoS, and set up its mathematical model, finding out a strategy. Then it explain Ant Colony algorithm, the basic principles and characteristics, processes, research in parameter setting. Go through domestic and foreign scholars on the improved Ant Colony algorithm and summary of operations, presenting the new strategy to better solve such a complex QoS routing optimization problem. Mainly in the following aspects:1. The idea of using parallel Ant Colony algorithm and Ant Colony Optimization into two independent subgroups separately.2. Two groups of pheromone of ant colony with different ways to ensure the convergence strengthen and prevent a better balance between the local optimal state.3. Using Ant Colony elitism, intervals as the elite of the cycle will be the best Ant and reward in order to achieve better results. Given the improved algorithm of parallel Ant Colony algorithm flow and implementation steps, and first used for TSP to verify its validity, then to QoS problems, the network topology is generated by random Waxman-Salama model experiment and further validate the improvement algorithm in QoS routing optimization of performance and function.Results show that the proposed algorithm can be used to identify optimal multicast tree, compared to Ant Colony algorithm, the indicator reached is satisfactory.Through this work shows that this algorithm in QoS multicast routing optimization problem is valid, get better results solving.While this algorithm is a general algorithm can provide references for other engineering fields, have a certain theoretical significance and applied value. |
