Abstract: The MinNLP based on minimum node and priority load is proposed aimed at limitation of the existing node mapping algorithm on the node load. The strategy uses node memory load as duty cycles parameter, chooses the minimum load and large residual resources of physical nodes to accept the mapping by calculating the load strength of the physical node, and uses the K shortest path algorithm to link the mapping between the nodes. The experimental results show that the mapping strategy based on minimum node and priority load and combined with K shortest path to construct logical bearer network node can obtain higher resource utilization, build success rate and link balance than the greedy method.

Key words: minimum node, priority load, logical bearer network

摘要： 针对现有节点映射算法在节点负载不均衡的问题，提出了最小节点负载优先的映射策略（MinNLP），该策略以节点内存负载作为判断节点忙闲度的参数，通过计算物理节点的负载强度，选择具有最小负载和较大剩余资源的物理节点来接受映射，并采用K短路径算法进行节点间的链路映射。实验结果表明，采用基于最小节点负载优先的映射策略并与K短路径思想相结合的方法来构建逻辑承载网比用贪心节点映射算法能得到更高的资源利用率、构建成功率和链路均衡度。

关键词: 最小节点, 负载优先, 逻辑承载网