Research On Self-Adaptive Clustering Routing Protocol In Wireless Sensor Networks For Meteorological Observation

In recent years, the rapid advancement of sensors and wireless communications has promoted the popularity of wireless sensor networks in an increasing number of areas. Sensor nodes in addition to have the ability to communicate, but also includes computing and storage capacity, which can form a topology network and accomplish the task of communication between nodes. In the traditional meteorological observation, the complexity of the monitoring environment adds constraints to the observation mode, and the cost of the traditional equipment is expensive. Therefore,the application of wireless sensor networks in the field of meteorology can get more meteorological data and provide better weather services. This paper do some research on improving the energy efficiency of the nodes, balancing the energy load and prolonging the lifetime of the meteorological observation system:Firstly, a cluster density adaptive clustering routing protocol - DAHC protocol is proposed. The protocol improves the single use of distributed or centralized clustering algorithm, and proposes the idea of mixed use. In the centralized clustering, the principle of LEACH-C protocol is used to select the optimal cluster head set, and then the next round cluster head is selected based on the information of the previous cluster head. At the same time, in order to balance the network energy load between cluster heads, an adjustment mechanism is proposed to make the cluster density adaptive for the case of uneven distributed clustering. Finally, experiments show that the method can effectively balance the energy consumption of nodes, prolong the time of observing meteorological data, and improve the effectiveness of energy.Secondly, the application of the protocol for homogeneous networks may not be suitable for heterogeneous networks because of the differences in the capabilities of the nodes in practical applications. Therefore, this paper introduces a number of high-level nodes in the network to form a two level energy heterogeneous sensor network.At the same time, the cluster head selection strategy is improved. The probability of the high-level nodes with more initial energy to be cluster head is larger than that of the ordinary nodes when the centralized election is used. The residual energy and location factors of nodes are considered in distributed election. The cluster density adjustment mechanism is used to balance the load between cluster heads. Finally,the experimental results show that the algorithm can make full use of the power of advanced nodes, prolong the life cycle and improve the throughput of the base station.