Research Of Energy-Efficient Routing Protocol Based On Dynamic Clustering For Wireless Sensor Networks

A Wireless Sensor Network (WSN) is generally composed of hundreds or thousands of wireless sensors which can sense and collect information from surrounding environment. The wireless sensors are typically deployed in remote or difficult to be accessed natural environment, such as battlefield, volcanos, glaciers of Arctic and Antarctic, or on the bridges, etc. They can assist human to monitor the sensitive information in certain places. However, the sensors in WSN are powered by battery which cannot be recharged after initial installation. This makes energy conservation the primary issue to be considered while designing the WSN system. Among all the energy consumption, most is consumed during data transmission. Therefore, an energy-efficient routing protocol will effectively relieve the pressure of limited power resources.Besides limited energy, as is subject to low cost and restricted size, there is also a limitation in memory and computing ability of the sensors. Therefore, besides energy conservation, low time and space complexity are also necessary. However, most existing routing algorithms only place emphasis on energy conservation, appropriate algorithm complexity were neglected. This dissertation takes advantage of the strict time synchronization among sensor nodes, designs the energy based time back-off mechanism to select cluster head and a dynamic cluster head pick method for member nodes. New algorithm achieves both energy conservation and low algorithm complexity.Firstly, during the cluster heads (CH) selection phase, energy was considered into the back-off time of the sensors who would like to become a CH. This mechanism makes the nodes that have more energy to be more likely to win the CH competition; new algorithm makes the nodes to broadcast the CH statement with a fixed radius to control the number of cluster heads.Secondly, in clustering phase, each non-CH node joins cluster by tradeoff between the distance from itself to the CHs and the distance from the CHs to the base station. Under the premise of preserving their own energy, the non-CH nodes join different CHs to save the energy of the CHs. This promotes the energy balance among global nodes, which in turn will prolong the network lifetimeSimulation experiments evaluate the algorithm by conducting the comparison in terms of CH distribution, the stability of the number of CHs, clustering performance and the lifetime. The results indicate that new algorithm improves by 33.4% and 18.2% compared with LEACH and HEED respectively in terms of network lifetime. The time complexity of new algorithm is O(1), which is lower than O(n) of HEED and many other classic algorithms. New algorithm achieves both the characteristic of energy conservation and low algorithm complexity.