HUNA: A Method of Hierarchical Unsupervised Network Alignment for IoT

摘要: With the advent of era Internet Things (IoT), a large number interconnected smart devices form huge network. The network can be abstracted as graph, and we propose to identify similar IoT in different networks by graph alignment. However, most methods rely on prelabeled cross-network node pairs such anchor links, which are difficult obtain due personal privacy security restrictions, especially IoT. In addition, existing entity alignment focus individual nodes but ignore tightly connected group structure network, is significant feature devices. this article, method hierarchical unsupervised (HUNA) deep learning approach. First, an based cycle adversarial (UNA), utilizes characteristics achieve under conditions. Second, further expand model carefully designing aggregation optimization module aggregate with closely related attributes structures into coarse-grained align nodes. Finally, evaluate HUNA real synthetic data sets. Experimental results show that improve accuracy 10% perform well terms parameter sensitivity.