Hybrid Deployment Algorithm Based On Different Virtual Force Models And Corresponding Parameter Analysis In Wireless Sensor Network

The dynamical and adaptive node deployment of a wireless sensor network is a research hotspot in the field of current deployment algorithms in special environments.Many existing deployment algorithms have a relatively single role,are only applicable to a small number of sensor nodes,and can not be easily deployed on large-scale nodes.Furthermore,they are not applicable to the deployment of threedimensional sensor networks.Recently,our group developed a dynamic deployment algorithm based on the dust plasma Yukawa potential model,which can be improved to a three-dimensional application.We try to use this algorithm to combine different virtual force algorithms and complement each other.It will speed up the convergence of the sensor deployment and improve the efficiency and performance of algorithm deployment.In this paper,we focus on the fusion and optimization of two different virtual force models and corresponding parameter analysis in a wireless sensor network application.We use an important performance index,namely Pair-Correlation Diversion(PCD),to analyze the similarity between the network grid and the perfect regular hexagonal grid.We combined one virtual force algorithm based on Yukawa potential model from dusty plasma(i.e.VAF_DP)and another virtual force algorithm based on Lennard-Jones potential(i.e.VAF_LJ).Although these two algorithms used the same second-order differential equation as the node's motion equation,they have different characteristics and can realize the complementary advantages of the two algorithms.In the research process,the research results of the VAF_LJ algorithms are summarized on the basis of the present achievements.And a new method based on Delaunay triangulation is used to accurately acquire the neighboring nodes of all nodes.Furthermore,the deployment grid is optimized.This technology is denoted as VAF_DJ algorithm.The different influence factors of algorithms in the VAF_DP algorithm are discussed and analyzed,and conclusions are drawn.As we know,the VAF_DP algorithm is suitable for large-scale node deployment solutions,and the VAF_DJ algorithm is suitable for medium-scale node deployment solutions.The final deployment grid based on VAF_DP algorithm is closer to the perfect regular hexagonal grid than that based on VAF_DJ algorithm.However,the deployment time of the VAF_DP algorithm is much longer than that of the VAF_DJ algorithm.The final deployment grid of VAF_DJ algorithm has better uniformity than the VAF_DP algorithm.After study of two kinds of virtual force algorithms,two kinds of algorithms are compared and analyzed.Based on the basic characteristics of the two algorithms,we believe that both algorithms can be effectively fused and applied to the dynamic deployment of large-scale wireless nodes.The VAF_DP algorithm adjusts the grid distribution as a whole,and the VAF_DJ algorithm continuously adjusts the grid distribution locally.Based on this,we proposed a fusion strategy based on both VAF_DP and VAF_DJ,which forming a hybrid deployment algorithm(i.e.VAF_Hybrid).Then the practicability and reliability of the algorithm fusion are proved from the convergence speed of the algorithm,the value of PCD,and the uniformity.In the end,we conducted a preliminary exploration of how to create a wireless sensor network experimental platform.