Firstly a maximum similarity distribution model is proposed. Secondly, a new definition, named virtual coordinate of a node which composes of the minimum hop counts to several special nodes, is introduced in instead of absolute physical coordinate of a node. Based on the theories, a distributed location-independent node scheduling scheme is proposed. The scheme consists of a coverage algorithm and a connection algorithm. The coverage algorithm takes advantage of the nodes' virtual coordinate to divide all nodes into several subsets. Without using location information, this scheduling scheme not only has the ability that sensor nodes in subsets are more uniformly distributed in the target region than other schemes, but also guarantees that all subsets are connective. The simulation results show that the scheme outperforms the randomized node scheduling scheme, on the coverage rate and network lifetime, as well as the number of external nodes when maintaining subsets to be connective.