论文部分内容阅读
针对无线Mesh网络网关节点和网络链路承载的负载不均问题,择优选择网关节点,并设计链路权重,构建以网络加权吞吐量为优化目标的资源分配模型.在构建的资源分配模型下,提出一种基于Q学习和差分进化的联合功率控制与信道分配算法(QDJPCA).该算法通过获取功率控制的反馈结果,采用基于多重变异和自适应交叉因子的差分进化算法进行信道分配;针对每次迭代产生的信道分配结果,采用基于状态聚类和状态修正的Q学习算法实现功率控制.NS-3仿真结果表明,QDJPCA能够有效求解所提资源分配模型,在优先保证网关负载均衡和高负载链路吞吐量性能的基础上提升网络整体性能.
Aiming at the problem of load imbalance between gateway nodes and network links in wireless Mesh networks, gateway nodes are chosen optimally and link weights are designed to construct a resource allocation model with network weighted throughput as the optimization objective.Under the resource allocation model, This paper proposes a joint power control and channel allocation algorithm based on Q learning and differential evolution (QDJPCA). This algorithm uses the differential evolution algorithm based on multiple mutation and adaptive crossover to obtain the feedback of power control. And the Q-learning algorithm based on state clustering and state correction is used to realize the power control.The simulation results of NS-3 show that QDJPCA can effectively solve the proposed resource allocation model, which can guarantee the gateway load balancing and high load Based on the throughput of the link to enhance the overall network performance.