Secure Data Aggregation Technique For Wireless Networks
Keywords:
Atmospheric Scattering Model, Dark Channel Prior, Color Attenuation Prior, Mean Square Error, Structural SimilarityAbstract
Due to restricted procedure power resources and energy resources, aggregation of data from multiple
sensing element nodes done at the aggregating node is sometimes accomplished by easy strategies like averaging. but
such aggregation is thought to be extremely liable to node compromising attacks. Since WSN are typically unattended
and while not tamper resistant hardware, they're extremely prone to such attacks. Thus, ascertaining trait of knowledge
and name of sensing element nodes is crucial for WSN. Because the performance of terribly low power processors
dramatically improves, future individual nodes are going to be capable of playing additional refined information
aggregation algorithms, therefore creating WSN less vulnerable. Repetitive filtering algorithms hold nice promise for
such a purpose. Such algorithms at the same time mixture information from multiple sources and supply trust assessment
of those sources, typically in a very type of corresponding weight factors allotted to information provided by every
supply. During this paper we tend to demonstrate that many existing repetitive filtering algorithms, whereas considerably
additional strong against collusion attacks than the easy averaging strategies, are nonetheless susceptive to a unique
refined collusion attack we tend to introduce. To deal with this security issue, we tend to propose AN improvement for
repetitive filtering techniques by providing AN initial approximation for such algorithms that makes them not solely
collusion strong, however conjointly additional correct and quicker joining.
