Optimization Method Of Low Power Topology Structure For Multi-node And Complex Sensor Network

Wisdom agriculture based on wireless sensor network promotes the development of agricultural information, due to the self-organizing and fault tolerance, multi-node and complex sensor network has become a hot topic. Throughout the research of domestic and foreign countries, the network is complex and sensitive to the environmental conditions, which contributes to the poor adaptability,low energy efficiency and transmission congestion, the real-time and accuracy of network are greatly influenced.Therefore, an optimization method of low-power topology structure for multi-node and complex sensor network is proposed in this paper, Moreover, the topology model with strong adaptability and robustness is established, the main contents of this thesis are as follows:1. Hybrid topology model of Layer cluster based on regular hexagon is studied for low node utilization, hexagonal deployment is adopted for outer ordinary sensor node, ordinary nodes and cluster head nodes are connected using the star structure, the inner cluster head nodes are classified according to the circular domain, the cluster head nodes of each grade are connected using the reticular structure, the cluster head nodes communicate with the base station by using the method of mixed hop routing based on critical distance, the perception efficiency and costs are improved.2. The dynamic cluster head selection method for wireless sensor network is proposed for energy consumption, the coverage redundancy nodes are selected as the first kind of cluster head nodes, after the death of first kind of cluster head nodes, then the survival time estimation algorithm is used to obtain the average remaining energy of network, the nodes of maximum energy ratio are selected as a new kind of cluster head nodes, it balances the energy of network in two phase with high adaptability and accuracy.3. An adaptive multi-path approach based on the improved leapfrog algorithm is proposed for transmission congestion, the path satisfaction model which takes the predictive congestion degrees, remaining energy and the minimum number of hops into consideration is established. The updating algorithm is discrete and learning factor is introduced during the local optimization process; during the global optimization, the introduction of multi-path routing makes it possible to achieve the optimal path. The optimal path is selected adaptively and the reliability of information is improved.According to simulation experiments, the energy efficiency of hybrid topology model is increased by about8.6%compared to quadrilateral topology model; the network lifetime of dynamic cluster head selection algorithm is respectively increased by50%and30%compared to the traditional LEACH and DEEC algorithms; the congestion degree of improved leapfrog algorithm is respectively decreased by14.7%and20.6%compared to the traditional ABEAR and AODV algorithms, The search accuracy of optimum path is up to97%. Experimental results show that, the energy efficiency is improved and network reliability is enhanced.