Introducing The Artificial Ant Colony Algorithm Of The Concept Of "ant King" And Its Application

Quality of Service(QoS)has been one of the research focuses in the Internet world.The QoS on network unicast and multicast requires finding best routes of transmission under various preset conditions such as bandwidth,delay and jitters.Thus,network routing optimization can be seen as a NPC multi-objective and multi-constraint math problem,for which traditional algorithms are not effective enough to get the optimal solutions in polynomial time.Ant colony optimization is an effective swarm intelligence optimization algorithm,which could solve NPC problems on the premise of ensuring QoS.Compared with other optimization algorithms,the ant colony optimization has advantages in processing parallelism,fast convergence speed,good robustness and wide application range,which is applicable to many types of practices and has great application value for optimization problems both under constrained and unconstrained conditions.We briefly introduce the structure and main contents of this thesis in Chapter 1,as well as the research importance of ant colony optimization.In Chapter 2,ant colony algorithms are discussed with respect to their basic definitions,principles,and improvements.In Chapter 3,we mainly discuss the origin and architecture about quality of service of computer networks.In Chapter 4,a new concept named "queen" is introduced to the algorithm,which can help in storing,sorting,and guidance for the process of finding optimal paths and finally make the group search not only more ordered,but also more efficient..In Chapter 5,the improved ant colony algorithm with a queen will be used for QoS multicast routing computation.In this process,each ant will deliver the path information and related pheromone concentration to the queen.Using a pheromone matrix about elite solutions,the queen will order the stored paths from short to long by comparing,abandon the long ones and save the short ones for the following ants to use,and dynamically update the pheromone on the paths.Then the pheromone matrix will update with latest findings.After many times of iterations,a new shortest path will be formed.The state transition rule of the ants is not only dependent on the influence of external environment of path pheromone,but also on the guidance of the queen.Because the queen always saves the best paths up to date,the ants will be able to decide next path of choice according to info from the queen,so as to alleviate the stagnation and local optimum phenomenon in traditional ant colony algorithms.In order to verify the validity of the ant colony algorithm with a queen,we use Matlab to simulate the QoS multicast route choosing problem in static computer networks.Firstly,the Salam topology algorithm is used to randomly generate a wired network for our research.Then we run the ant colony algorithms with and without a queen respectively,and compare their optimized results.The experimental data show that,for solving practical problems like QoS multicast routing,the searching performance of ant colony optimization with a queen is obviously superior to the basic ant colony algorithm.