基于改进遗传算法的多约束QoS路由算法研究

遗传算法良好的全局搜索能力使其被广泛地应用于网络中多约束QoS路由寻址,并取得了较好的成果。然而大部分应用于多约束QoS寻址的改进遗传算法存在无法有效利用网络资源使得网络拥塞、网络过早收敛陷入局部最优解,以及过慢结束的缺点。针对上述问题,对传统遗传算法中的适应度函数和遗传算子做出相应的改进,并通过实验验证提出改进遗传算法。最终,仿真实验表明该算法是可行的,并能在一定程度上克服传统遗传算法的缺点,提高收敛速度,有效避免过早收敛,提高QoS满意率。

Multiple constrained QoS routing algorithm based on improved genetic algorithm

Genetic algorithms is widely used in multi-constrained QoS routing addressing because of its ability of global search, and the same time we have got good results.However, most of them applied to multi-constrained QoS addressing has the disadvantages of not effectively use the network resources which make the network congestion,and easily fall into local optimal solution, as well as the convergence of best solution is slower.In response to these problems,we improve the traditional genetic algorithm fitness function and the genetic operators in this paper,and the experiments were been done to verify it.And finaly,the simulation results show that the algorithm is feasible,and the defect of traditional genetic algorithm can be overcome to a certain extent,hence improve the convergence speed, and effectively avoid the premature convergence and improve QoS satisfaction rate.