In practical applications, Wireless Sensor Networks (WSNs) generate massive data streams with the dual attributes in geography and optimization domain. Energy source of sensor nodes in WSN is usually limited; Data streams transmission is known to be the largest consumer of energy in WSN. Therefore, reduce the total data transmission and maximizing energy efficiency is the major challenge in WSN. In addition, sensor data points usually have attributes in the optimization and the geography domain. In most existing clustering algorithms, the similarity measurement mainly takes the non-geometric attributes into consideration.