【作者】 李超峰；

【导师】 胡燕；

【作者基本信息】 武汉理工大学 ， 计算机科学与技术， 2011， 硕士

【摘要】 随着时代的发展和科学技术的推动,通信网络也不断的向前发展。在经历了PDH和SDH传送网络之后,通信网正向着PTN传送网的发展。PTN传送网是以分组IP为内核的传送网络,是下一代的传送网络。为了管理PTN传送网,PTN网络管理系统是必不可少的。在PTN网络中,所有的业务都承载在PW上,在确定了业务的源端点和目的端点后,根据相应的QoS约束寻找合适的PW称为寻找路由。由于网络和相关业务的复杂性,自动路由就成为必然。而寻找带QoS约束的路由问题是一个NP-难问题,在解决此类问题时,常规的方法难以满足要求,解决该类问题的方法一般是采用智能启发式算法,如遗传算法,蚁群算法等。本文主要研究了遗传算法和蚁群算法,在分析了遗传和蚁群算法的优缺点之后,结合两者的优点提出了一种基于蚁群的混合算法。并将该算法用于解决QoS路由问题,具体所做的工作包括如下几个方面：1.介绍了课题背景,PTN技术的发展现状和PTN网管的相关情况,并详细的介绍了QoS路由模型。2.分析了该PTN网管系统的架构及系统中的关键技术,介绍了在该系统中业务的创建流程。详细说明了自动路由和人工路由的优缺点,提出了自动的路由的必要性,以自动路由为本文的研究点。3.在解决QoS自动路由时,分析了遗传算法和蚁群算法的原理,流程以及在QoS路由问题中的应用。在深刻理解算法的优缺点基础上,提出了一种以蚁群算法为基础的混合算法,该算法先利用遗传算法的全局寻优能力和快速性,生成初始解,然后利用部分解来初始化蚁群信息素,并利用蚁群算法生成最终解。4.利用实验仿真来验证算法的有效性,实验仿真表明该算法在求精方面优于遗传和蚁群算法,在时间性能上,优于蚁群算法,该算法是有效的。并将该算法应用于实际的PTN网管中,用于创建业务时自动路由的寻找。根据实验仿真和最后在实际应用中表明,该混合算法比基本遗传算法和蚁群算法有着更良好的效果,是一种效果良好的算法。更多还原

【Abstract】 With the development of the times and promotion of science and technology, the communication networks continue to move forward. Experienced in the PDH and SDH transmission network, the communication network is toward the development of transmission network of PTN. PTN transmission network based on IP as its core, is the next generation of transmission network. In order to manage PTN transmission network, network management system is essential.In PTN network, all business is carrying on the PW, after determined the source endpoint and purpose endpoint of the business, according to the corresponding authored QoS constraint find suitable PW called looking for routing. Due to the complexity of the network and related business, automatic routing is inevitable. In search of the routing problem with QoS constraint is a NP-hard problem. to solve such problems conventional method can not satisfy the requirements of this problem and solving methods are generally using intelligent heuristic algorithm, such as genetic algorithm, the ant colony algorithm, etc.This paper mainly studies the genetic algorithm and the ant colony algorithm genetic,after analyzing the advantages and disadvantages of the two algorithms, then the proposed hybrid algorithm combining with the advantages of both which based on ant colony. And the algorithm is used to solve a QoS routing problem, specific work done including the following aspects:1. Introduced the subject of background, and PTN technology development status and related network management, then detial the QoS routing model.2. Analyzes the structure and framework of PTN element management system and the key technology which are introduced in the creation of the system of business process. Detail the automatic routing and artificially routing advantages and disadvantages, puts forward the necessity of automatic routing for this paper, and take automatic routing as the research points.3. When in solving the QoS automatic routing, analyzed the genetic algorithm and the principle of ant colony algorithm, process and the application in the QoS routing problem. after profound understanding its advantages and disadvantages of the two algorithms, was presented based on ant colony algorithm based on hybrid algorithm, this algorithm by using the genetic algorithm first the global optimization ability and quickness, initial solution, then uses part of the solution to initialize ant colony pheromones, and use the ant colony algorithm generated eventually solutions.4. Use experimental simulation to verify the efficiency of the algorithm, experimental results show that this algorithm in refinement aspects due to genetic and ant colony algorithm. in time performance, better than ant colony algorithm, this algorithm is effective. And the algorithm was applied to the actual PTN net, used to create business automatic routing search.According to the simulation results and in practical application shows that the hybrid algorithm has better effect than the basic genetic algorithm and the ant colony algorithm, it is a kind of good method.更多还原