The function of neural network of human beings is quite similar to that of the sensor network, which is a cluster of firmly related individual units performing a special function. According to the architecture of neural network, a dynamic-cluster energy-aware routing algorithm (DCEAR) has been proposed in this paper. The nodes on the network have adopted an active routing based on event awaking. Firstly, the event data dynamically elects a temporary clusterheader. Then the temporary cluster-header handles all the data it receives, and the handled data reflects the general instance of the region. In this way it has decreased the information that has to be forwarded and has reduced the energy consumption of the entire network. The backbone of the network is originated by genetic-tree algorithm. It will timely update the trunk of the genetic-tree in order to trade off the energy consumption of the whole network. So it has prolonged the lifetime of the network greatly. The result of our simulation indicates that it is a quite effective energy saving routing algorithm.