Research On Energy-Efficient Routing Strategy For Tree Wireless Sensor Networks

With the scale increasing and hardware devices updating, the system of wireless sensor networks becomes more and more complex. Restricted by the size, cost and resources of wireless sensor nodes, the memory and computer power are limited. Typically, nodes are battery powered and the energy cannot be supplied unlimited, so many researchers have recently committed to prolonging the network lifetime through controlling the network topology, developing new route strategy for data forwarding and balancing the energy consumption of nodes. According to previous work and based on tree network we propose an energy-efficient routing strategy.Firstly, we need build a tree network by self-organization. During the network self-organization process, we divide the self-organization strategies into two kinds:active self-organization and passive self-organization, according to the cluster nodes’selection of non-network nodes. Non-network node selects the best cluster node from all available cluster nodes and the selection of best cluster node balanced the hop, residual energy, child node number and distance between nodes. When non-network node joins the network success and becomes a network node it will do the function of network node.Secondly, after the tree network is built all sensor nodes begin to collect data and transmit the data to their cluster nodes. Cluster nodes cluster all received data and then transmit them to their cluster nodes. Through the cluster and transmit one layer by one layer all collected data can be received by root node. During the network working process cluster nodes consume more energy. And if the cluster node closer to the root node it will consume more energy. So for balancing the energy consumption and prolonging the network lifetime we adjust the topology dynamic. If the energy of a cluster node reduces to R% we need to delete the farthest child node to balance the energy consumption. R% is based on the residual energy of last topology changing, which means the cluster node adds or deletes a child node. When the link is failure or the link quality is very low we also need to adjust the network topology dynamically. During the network part re-organization process, the link cycle must be avoided and we give a link cycle check method in this project.Finally, we do some simulation experiments based on NS2 and actual experiments based on STM32W108 ZigBee chips. The experiments results show that the new strategy can construct a reliability tree network quickly and the passive self-organization strategy is better. With the network scale increasing the self-organization time, average hops and packet loss ratio won’t increase more. The new strategy can balance the energy consumption of nodes and prolong the network lifetime.