基于队列稳定性的联合资源优化算法

Joint resource optimization algorithm based on queue stability

为了解决无线虚拟化网络中的多种资源动态分配问题,保证系统队列稳定性的前提下提高网络资源利用率,提出一种基于队列稳定性的联合资源优化算法。首先,该算法考虑了每个虚拟运营商的缓存资源限制,对不同的虚拟运营商采取不同的定价机制,将多种虚拟资源分配作为一个联合优化问题,以最大化虚拟运营商收益为目标建立效用函数。然后,运用Lyapunov随机优化方法,根据当前系统用户的数据积压量设计了一种分布式调度算法,对虚拟网络收益的最优性与系统队列的稳定性进行权衡控制。最后,利用Lagrange理论对模型进行迭代求解。仿真结果表明,所提优化方案能够保证系统稳定的前提下提高虚拟网络总收益。

To achieve the dynamic allocation of multiple resources and improve the resource utilization with ensuring the queue sta-bility, the paper proposes a joint resource optimization algorithm based on queue stability in wireless virtualized networks. Consider-ing the cache resource constraints and different pricing mechanisms for each virtual operator, the algorithm allocates multiple virtual resources as a joint optimization problem, and establishes a utility function aiming at maximizing the profit of virtual operators. Sec-ondly, based on the data volume, a distributed scheduling algorithm is designed to balance the optimality of the virtual network rev-enue and the stability of the system queue through Lyapunov stochastic optimization. Finally, the Lagrange duality theory is used to solve the model. The simulation results show that the proposed method can effectively improve the total average revenue of virtual network while guaranteeing the queue stability.