WMN中基于网关饥饿度的部署算法优化

Placement algorithm optimization in wireless mesh networks based on gateways' hungry-value

研究满足QoS约束条件的网关负载均衡部署优化问题,定义网关饥饿度衡量网关负载均衡性,并提出网关部署的饥饿算法,在为每一簇分配网络节点时,都尽量使其簇头(网关)饥饿度最大程度接近网络总的平均值,最终实现网关间负载均衡,同时满足QoS约束。仿真实验结果表明:饥饿算法得到的网关数量与其他传统算法得到的结果非常接近,甚至更优;而在网关负载均衡方面,饥饿算法优势较明显,与Greedy_Partition算法相比,网关饥饿度样本标准方差约减少54%。

The problem of gateways deployment was addressed to achieve the goal of load balance in gateway placement with QoS requirements being satisfied, the gateways＇ hungry-value was defined to measure the load-balance of gateways, and a hungry algorithm was presented for network clustering. When network nodes was assigned for each cluster, the hungry value of gateway should become close to their average of hungry value as much as possible and achieve load balance placement of gateways in the end. At the same time, it always meets the QoS constraints during the entire clustering process. The results show that the number of gateways generated by the hungry algorithm is nearly equal to those from other gateway placement algorithms, and as far as the load balance of gateways is concerned, the hungry algorithm performs much better than the others. Specially, compared with GreedyPartition algorithm, the hungry algorithm improves the load balance of gateways with the standard deviation of the gateways＇ hungry value decreased by 54%.