Abstract: Wireless sensor network (WSN) nodes can't gain durable energy supply. Therefore, how to use the limited energy efficiently to prolong the lifetime of nodes becomes an important research problem. Proposes a filter-based uniform algorithm (FBUA) for optimizing Top-k query in WSN. Its basic idea is to provide a filter for each and every node. The filter is essentially a node value fetch interval determined by sample value. When the node gets some new data, it will decide whether to update it in accordance to the filtering rule. Thus, a lot of redundant communications can be shielded to save the energy of node and prolong its lifetime. FbUA can get exact answers to some complex queries in WSN, such as Top-k and k-NN. The experimental results showed that in most cases FbUA could reduce bandwidth consumption by 50%-70%.