Design And Analysis Of Dynamic WSN Data Acquisition Protocol Based On Clustering

In recent years,wireless sensor networks(WSNs)have attracted a great deal of attention due to its versatility and have been widely used in environmental monitoring,health monitoring,transportation,underground pipelines,construction and military applications.Most applications require sensors to periodically sense the environment and then send the data to a remote central unit(e.g.,sink)for processing via multi-hop paths.Typically,sensor nodes are randomly deployed in harsh,inaccessible environments and rely solely on battery power,which makes it virtually impossible to be maintained once deployed.As we can see,energy is one of the most important factors that restrict the normal operation of the network.Therefore,collecting data in the most energy-saving manner is crucial for the life of the WSN.However,to successfully design and deploy a WSN,many problems need to be addressed,such as network deployment strategy,energy efficiency,routing,data aggregation delay,bandwidth utilization and so on.In addition,another important issue is the dynamic nature of the network.Existing data collection protocols generally set the nodes to be stationary after deployment,while ignoring the situation that the node death(energy depletion or failure),the new node to join,the node movement,etc.,which makes it that the design of the protocol has a very large limitation.In order to solve the above problems,this paper mainly focuses on the two aspects of energy consumption and dynamic of the network,and studies a cluster-based data collection protocol in dynamic WSNs.The specific research contents are as follows.(1)A review of the current data collection protocols is presented.For the data aggregation protocols,we propose seven kinds of classification methods according to the network topology,interference model,security,number of channels,algorithm execution mode,driving method and protocol stack implementation level,respectively,and mainly analyze the first four classification methods in detail.(2)The paper mainly classifies the clustering routing protocols from three aspects: the implementation of the clustering algorithm,the number of cluster layers and the applicable range of the clustering algorithm,and points out the characteristics of each category.(3)For dynamic WSNs,a novel clustering routing protocol CRPD(clustering routing protocol for dynamic networks)was designed to implement data aggregation.The protocol is divided into four phases,including five algorithms,and each algorithm is analyzed in detail.The basic idea is to periodically update the network topology,select the node with larger degree and higher residual energy as the cluster head to be responsible for data aggregation and transmission.With the nodes moving,joining and choosing the optimal clustering radius,the energy load of the whole network can be evenly distributed to each sensor node,which can significantly prolong the network lifetime.In this protocol,the use of clustering and routing techniques helps reduce energy consumption and improve energy efficiency.Experiments show that CRPD protocol is more energy-efficient than existing protocols.