Research Of Dynamic Deployment In Wireless Sensor Networks Based On Dusty Plasma Model

Wireless sensor networks(WSN) is a hot topic in current information field. It enables signal acquisition, processing and transmission under special circumstances,which is a new information acquisition and processing technology and has been widely used in real life.This article focuses on the research of adaptive dynamic deployment algorithm for wireless sensor network. The dynamic deployment of sensor networks solves the problem of uneven nodes distribution after randomly deploying and improves the coverage and connectivity of monitoring area, making the research of adaptive algorithm for wireless sensor deployments become a hot spot. The virtual force algorithm, inspired by the interaction among physical particles, is one of the most important methods to solve large scale dynamic deployment of sensor nodes.In this paper, we apply dusty plasma crystallization model in plasma physics to the dynamic deployment of wireless sensor networks, and proposes a optimal virtual force algorithm based on Yukawa potential. Seeing a sensor node as a dust particle,we use molecular dynamics methods to simulate the forces among the particles,making the wireless sensor networks achieve an effective and large-scale dynamic deployment. We detailedly describes the principles of the algorithm and its involved physical background. In the process of verifying the effectiveness of the algorithm,we introduces and improves a novel performance metric called pair-correlation diversion(PCD) to analyze our algorithm performance. Through simulation and comparison, it shows that our algorithm has good convergence speed and deployment performance.In our research, we introduce and improve an node self-deployment method based on virtual exchange force among particles. We simply describe the algorithm principle and then optimize the shielding rules, in which a new way based on Delaunay triangulation is adopted to accurate obtain the adjacent node of all nodes.Through simulation and comparison, we discuss our algorithm’s convergence speedand deployment effect and get better deployment performance. Finally, we detailedly investigate the influencing factors of the algorithm and obtain the optimal range of parameters, which provides the basis for selecting the input parameters in real experiments.