Research On Fusion Algorithm Based On Linear Counting In Wireless Sensor Networks

Recent advances in wireless communications and electronics have enabled the development of low-cost,low-power,multi-functional sensor nodes.These sensor nodes,which consist of sensing,data processing,and communicating components,are basic components to build wireless sensor networks(WSNs).There have been a wide range of potential applications for WSNs including military,environmental monitoring,health care,transportation management,commercial applications and so on.Data fusion technique is one of the key technologies in wireless sensor network applications.One of the most common data fusion strategies is the tree-based aggregation.However,sensor nodes often have a poor communication quality due to resource limitations,packets transmitted between two sensor nodes are often dropped.The fact makes the tree-based fusion to be poor in the quality of the fusion results.This paper use the multi-path-based fusion.It enhances the robustness of sensor data fusion by having each node broadcasts its reading or partial fusion to multiple neighbors.And introduces a data fusion algorithm based on linear counting to solve the duplicate counting problems,and its main features are as follows:1.The proposed algorithm introduces a linear counting,extends the linear counting technique to represent sensor reading values,and uses a random number instead of the hash function.It helps save node energy consumption.Then linear counting technique to estimate the result of fusion,that have high accuracy.2.The algorithm designs a sketch with characteristics of small storage space and duplicate-insensitiveness.The fusion of multiple data duplicate does not change the final value.It helps solve the problem of overcounting.3.The algorithm introduced a pair of parameters.Through setting different parameters,it can obtain an integration value whose accuracy and storage space overhead can get a compromise.The integration value has high accuracy and low variance.The validity and the advantages of the algorithm is verified by experimental validation and the comparison of AATSD algorithm in the integration of the relative error and storage space overhead and integration time.