Based on the fact that most distance-free algorithms can improve anchor node ratio to improve the precision of wireless sensor networks node localization

an extreme learning machine（ELM） localization algorithm introducing virtual nodes for wireless sensor networks is proposed. Through introducing virtual nodes

looking for suitable unknown nodes to upgrade to sub-anchor nodes

so as to increase proportion of anchor nodes and improve positioning precision.Applying ELM to node localization can effectively improve speed and precision of localization and its powerful generalization performance can provide new ideas for localization of nodes in wireless sensor networks.The simulation result indicates that introducing ELM localization algorithm and virtual nodes can improve positioning precision effectively.