Optimizing Lifetime Of Wireless Sensor Networks

Wireless Sensor Networks (short for WSN) can get objective physical information; it has a wide range of applications, mainly used on military, national defense, industry and agriculture controlling and so on. WSN has involved many subjects and now has become a hot researching spot in the IT area.Nodes in WSN usually have constrained energy, and because network lifetime is closely related with energy consumption, once a node drains out of its energy, the whole network will not run properly. So prolonging the lifetime of the whole network becomes a challenge in the WSN researching area. Based on the study of current routing algorithms, this paper introduces a novel routing algorithm called the Energy Prediction and Ant Colony Optimization Routing(short for EPACOR), which uses energy aware Ant Colony Systems (short for ACS) in the process of routing establishment, and also uses Reinforcement Learning theory in energy prediction. Based on EPACOR, another novel algorithm is proposed, which is called Enhanced Ant Colony Optimization Routing (short for EACOR).In the EPACOR, when a node needs to deliver data to the sink, ACS are used to establish the route with optimal or sub-optimal power consumption, and meanwhile, reinforcement learning is embedded to predict the energy consumption of neighboring nodes when the node chooses a neighboring node added to the route. Then based on EPACOR we have improved the algorithm, before choosing any path gets from the ACS, we will pre-deal with the paths focusing on protecting the key nodes, so they won't be drained out of energy .It also improves the description of the route path, so it can reasonably describe weather the path is good or not. We call this routing algorithm the EACOR. Worth to be pointed out, because of using energy prediction based on Reinforcement Learning, sensor nodes needn' t to exchange information about node' s residual energy. It helps the network decreasing the energy consumption for controlling signaling.The research results of this paper have important application values in intelligent transportation, environment detection, ecological protection, nature disaster forewarn and such kind of areas.