Privacy Preservation Algorithm For Data Aggregation Based On MIN/MAX In WSN

Nodes in Wireless Sensor Networks (WSNs) collect data. After data aggregation, the resultsare sent to the sink node, and analyzed by user to obtain useful information. Data aggregationtechniques are effective measures to reduce traffic of node; privacy-preserving of node data is thebasic requirement of WSNs. So it is essential to do research on privacy-preserving aggregationtechnology in WSNs which supports data aggregation. MIN/MAX function has a wide range ofapplication scenarios, It is significant for the specific condition and research area of min/maxinformation requirement，so privacy-preserving aggregation for min/max computation in WSNs isthe main research point of this thesis.In this paper, a mount of research is done on the data aggregation techniques. On this basis, asuitable MIN/MAX data aggregation privacy-preserving algorithms GRIPDA is proposed. GRIPDAalgorithm constructs a staggered black and white clustered structure. The cluster head generates arandom interval sent to the leaf nodes in the cluster, the leaf node calculates its own data with therandom data on a certain range of operations, and the computed values are transmitted in paths ofdifferent colors. The value replaces the source data value of sensor nodes, which achieves the dataprivacy-preserving during transmission.TinyOS simulation based on ZigBee protocols isaccomplished with TOSSIM on the platform of TinyOS system. Meanwhile, transmissionprocedures among sensor nodes are described and transmission costs under both beacon mode andnon-beacon mode are analyzed as well in this dissertation. The GRIPDA algorithm is simulated onTinyOS platform with the tool of TOSSIM. The algorithm performance is analyzed in detail fromfour perspectives of the privacy-preserving, data traffic. The simulation result shows that, on thepremise of ensuring data privacy-preserving in WSNs, GRIPDA algorithm is a low computationalalgorithm, and can effectively reduce the data traffic in the network. To some extent, the algorithmeliminates the energy losses of sensor networks, and extends the life cycle of the entire sensornetwork.