The virtual network embedding/mapping problem is a core issue of network virtualization. It is concerned mainly with how to map virtual network requests to the substrate network efficiently. There are two steps in this problem: node mapping and link mapping. Current studies mainly focus on developing heuristic algorithms

since both steps are computationally intractable. In this paper

we propose a new algorithm based on the proximity principle

which considers the distance factor besides the capacity factor in the node mapping step. Thus

the two steps of the embedding problem can be better integrated and the substrate network resource can be used more efficiently. Simulation results show that the new algorithm greatly enhances the performance of the revenue/cost (

R

/

C

) ratio

acceptance ratio

and runtime of the embedding problem.