During the routing, Data is collected from the all sensor nodes, in result, the sensor nodes located closer to sink node have to receive and transmit not only their own data but also data from other sensor nodes, which are further sent to sink. It leads to the fact that the nearest sensor nodes fail much earlier than remote ones, In the long run, traffic from all the sensor nodes is going through one sensor node that is nearest to sink and the former disrupt the work of the rest of WSN. From the point of view of energy saving, big WSNs with only one sink cannot consume resources effectively. To solve this problem it is necessary to divide the WSN into clusters. The new algorithms proposed with the name of Predictive packet reception with energy (PPRE) with the focus of cluster formation and data transfer used in wireless sensor networks by using soft computing techniques. The proposed algorithm would improve the quality of data transfer in a granted way through by increasing the overall packet delivery fraction, throughput, stability of whole networks. The proposed scheme is simulated over NS 2.34 and MATLAB.