基于优化遗传算法的负载均衡策略研究

移动互联网和云计算技术的飞速发展,孕育了一种新的移动服务框架——移动云计算(mobile cloud computing,MCC)。传统遗传算法(standard genetic algorithm,SGA)作为一种元启发式算法,在解决移动云环境中负载均衡问题时存在着收敛速度慢、易于陷入局部收敛的问题。提出了一种优化的负载均衡算法,采用基于多目标的适应度函数,自适应的交叉和变异算子,更好地适应了移动云环境中资源的异质性特征和资源部署的实时性需求。采用云仿真器Cloudsim进行仿真,结果表明改进算法在负载均衡度和资源部署时间方面均优于传统遗传算法。

Load balancing strategy based on improved genetic algorithm

The fast development of mobile internet and cloud computing introduces a new mobile service architecture- Mobile Cloud Computing （MCC）. We propose a new load balancing strategy based on the improved genetic algorithm to make better utilization of the resources in the mobile cloud environment. This algorithm applies multi-target fitness function and self-adaptive crossover and mutation operator, and solves the problem about slow convergence and falling into suboptimal convergence solution of the Standard Genetic Algorithm （SGA）. The Cloudsim platform is used for simulation and the result shows the optimized algorithm can suit the mobile cloud environment well and get better load balancing solution.