摘要
应用遗传算法优化小区无线网络规划,对传统遗传算法进行改进,使用Hata模型的COST231扩展作为电波传播模型,来计算个体的适应度,在进化操作中采用倒序单点交叉和二次反转变异因子。改进的遗传算法具有针对性,加快了种群的收敛并且避免了早熟,得出基站布局的最优方案,利用最少的基站数实现规定的服务质量,对小区移动网络中的基站位置进行优化。仿真结果证明,与采用传统的GOAT工具箱相比,改进后的算法更逼近Pareto域(即最优解集),还可以预测小区每一点处的场强大小,优化后的基站布局使覆盖率达到了90%以上。
The genetic algorithm is used to optimize cell network planning, which is improved in the tradi- tional genetic algorithm by expanding the model CONST231 Hata to calculate the fitness of in- dividual. Reverse single-point cross of order and the second reversal mutation factors are em- ployed in the evolutionary operations. The improved algorithm is able to accelerate the conver- gence and avoid premature and achieve quality of provided service with the minimum number of base stations, so it can optimize base-station-location in mobile networks. The simulation result shows that the improved algorithm can approach Pareto domain（the optimal solution set）as near as possible in comparison with the traditional toolbox of GOAT, and predict the field strength of each test points in this area. After the optimization the coverage with base-station is more than 90%.
基金
广西信息与通讯技术重点实验室基金(No.10905)
广西研究生教育创新计划资助项目(No.2010105950809M14)
关键词
电波传播模型
遗传算法
倒序单点交叉
二次反转变异
Radio Propagation Model
Genetic Algorithm
Reverse Single-Point Cross
Second Reverse Mutation
