It is a hard work for the traditional k-means algorithm to perform data clustering in a large, dynamic distributed wireless sensor networks. In this paper, we propose a distributed k-means clustering algorithm, in which the distributed clustering is performed at each sensor with the collaboration of its neighboring sensors. To extract the important features and improve the clustering results, the attribute-weight-entropy regularization technique is used in the proposed clustering method. Experiments on synthetic datasets have shown the good performance of the proposed algorithms.