This paper focuses on the selection of cluster head using Neural Networks for optimizing the network lifetime which could be used for energy efficient routing in wireless Sensor Networks. In cluster-based routing, special nodes called cluster heads form a wireless backbone to the sink. Each cluster heads collects data from the sensors belonging to its cluster and forwards it to the sink. In heterogeneous networks, cluster heads have powerful energy devices in contrast to homogeneous networks where all nodes have uniform and limited resource energy. It is essential to avoid quick depletion of cluster heads. Hence, the cluster head role rotates, i.e., each node works as a cluster head for a limited period of time. Energy saving in these approaches can be obtained by cluster formation, cluster-head election, data aggregation at the cluster-head nodes to reduce data redundancy and thus save energy. The first part of this paper discuss the methods for clustering to improve energy efficiency of homogeneous WSN and also proposes a Neural Network (NN) based "Winner takes all" algorithm for cluster head selection for WSN. The simulation results show improved performance of NN based optimization in terms of total energy dissipation and number of alive nodes of the network system over LEACH, Static Clustering and direct methods. © EuroJournals Publishing, Inc. 2012.