One of the most critical issues in designing wireless sensor network is to minimize the energy consumption and increase longevity. In Wireless Sensor Network, data fusion can reduce data redundancy among sensed data and optimal sensor routing algorithm can provide strategy for data gathering with as minimum energy as possible. We combined data fusion and clustering method together to reduce the energy. In this paper we propose K-means algorithm for clustering. We design a novel routing algorithm called an Adaptive algorithm(AA) for energy efficient data gathering in sensor network that optimizes data transmission cost and the data fusion cost. Simulation result is given for the proposed protocol with K-means Algorithm. Simulation result shows that AA achieves better performance than MFST.