Data Fusion Algorithm Based On Delay-aware Clustering In Wireless Sensor Networks

Wireless sensor networks(WSNs)is a kind of Ad Hoc networks connected by massive wireless sensor nodes(SNs),which transmit data to each other by wireless way.It has the advantages of rapid and convenient deployment,self-organization,data centric,strong concealment and high fault tolerance.In WSNs,energy consumption is one of the most important hinders that limit the development of WSNs.Data Fusion technology can reduce the amount of transmission data by removing redundant information and form high-quality fusion data to sink.So it is deemed to be one of the effective ways to solve the crisis of energy consumption in wireless sensor networks.Although data fusion can reduce the energy consumption of SNs,it often increases the delay of the networks,making the network delay and energy consumption become two contradictory indicators.Many sensor data requires real-time reporting in order to reflect the value of the data,such as earthquake monitoring,target tracking and fire alarm,this needs to adopt a WSNs algorithm which can save network energy and prolong the SNs' life,and can also guarantee the network delay.Therefore,how to build a delay-aware and energy efficient wireless sensor network model based on data fusion is a very promising research direction.First,we present the characteristics of the existing data fusion models in WSNs.Then according to the different algorithm used in the integration process,the WSNs data fusion technology is divided into four categories: data fusion based on statistics,data fusion based on artificial intelligence,data fusion based on information theory and data fusion based topological.Then,the existing problems and challenges of data fusion technology in WSNs are summarized and analyzed,and future research hotspots and directions including auto-convergence,fusion reliability,fusion security and fusion performance evaluation are proposed.Aiming at the problem of tradeoff between network energy consumption and delay in WSNs,and considering the node energy consumption optimization and load balancing at the same time,a data fusion algorithm based on hybrid delay-aware clustering(HDC)in WSNs is proposed.The network model of HDC based on single-cluster structure and multi-cluster structure is build at first.And the network performance of two structures under different network scale and data fusion rate is compared.Secondly,in order to improve the network performance and avoid the death of cluster head node,the clustering and reorganizing phase are optimized.HDC uses the distance and the residual energy as two reference factors to design an effective energy-efficient clustering process and a dynamic cluster head re-selection process,further reducing the total energy consumption of data transmission between nodes and rotating the cluster head nodes to achieve energy consumption load balancing.Finally,HDC is verified by Matlab7.1.The simulation fully considers the network environment under different network scales and different data fusion rates,compares HDC with the other three kinds of WSNs delay-aware models from three aspects of network delay,network energy consumption and network life cycle.From the simulation results,compared with the traditional delay-aware WSNs model,HDC algorithm has advantages in reducing network delay,saving network energy and prolonging the SNs' life cycle.