Research On Layered Adaptive Compression Data Collection And Privacy Preservation

With the continous development of smart earth and city,the sensory data fall across the new era of explosive growth for the continuous deployment of wireless sensing devices.Wireless sensor network(WSN)is taken as one main source of big sensory data,which is being widely used in intelligent transportation,environmental monitoring,industrial production and other fields.However,the energy of sensor nodes is very limited,and they usually cannot be charged in time.At the same time,due to the complex deployment environment of WSN,data collection schemes are still facing security threats.Conventional WSN can no longer meet the requirements of processing and managing massive data.Therefore,how to collect data efficiently under the premise of ensuring data security to achieve the goal of prolonging the network life cycle has become a key problem to be solved urgently.Aiming at the problems in the current data collection scheme,such as mininizing spatio-temporal correlation,sparse dictionary design and data privacy preservation,the innovative contributions of this paper include the following three aspects:1)Layered compression mechanism for efficient data collection of sensory data.Aiming at the problems of mining data correlation and improving compression rate(or reconstruction effect)in current data collection mechanisms,a layered compression mechanism for efficient data collection of sensory data(LCS-EDC)is designed to support the exploration of spatio-temporal correlations.Then,a specific projection method is constructed respectively for exploring the temporal correlation in sensory nodes,spatial correlation in cluster heads and spatial correlation in processing nodes.At the same time,the detailed solving method is developed to recover original data and achieve the approximate data collection in sink node.Finally,simulation results indicate that the proposed layered compression mechanism has better recovery performance as compared with traditional clustered compression mechanisms(i.e.,achieving efficient data collection with high quality).2)Layered adaptive compression design for efficient data collection.In order to improve the performance of efficient sensing data collection mechanism based on layered compression,and further mine the temporal correlation of sensing nodes,the spatial correlation of cluster head nodes(within clusters)and the spatial correlation of processing nodes(between clusters),this part makes an extended reaserch.By training the sparse dictionary,the scheme can adapt to different types of data,obtain better sparse representation,and further improve the accuracy of data reconstruction.Finally,through the simulation of real data,the results show that the proposed mechanism can achieve high-efficiency and high-quality data acquisition and better reconstruction performance.3)Spatio-temporal correlation based privacy preservation data collection mechanism.In view of the data collection mechanisms still face security threats in WSN,this mechanism designs a light encryption method,which guarantees that the privacy of data is not attacked by external eavesdroppers and internal curiosity.Moreover,this encryption method does not destroy the correlation of data,and is conducive to encrypting sampled data and performing decryption and decompression operations at the same time.In addition,the mechanism independently sampled and encrypted at each sensory node(including cluster head nodes),reconstructed and decrypted jointly at the link node,fully exploited the spatio-temporal correlation of data,and the relative reconstruction error was smaller.Finally,the simulation results show that the mechanism can not only collect data efficiently,but also ensure the security of its process.