AN EFFORT TO DETERMINE CREDIBILITY AND STATUS OF SENSOR NODES

Abstract

Iterative filtering algorithms hold great promise for this type of purpose. Because of limited computational power and sources, aggregation of information from multiple sensor nodes done in the aggregating node is generally accomplished by simple methods for example averaging. Within this paper we show several existing iterative filtering algorithms, while considerably better quality against collusion attacks compared to simple averaging methods, are nonetheless susceptive to some novel sophisticated collusion attack we introduce. However such aggregation is proven to be highly susceptible to node compromising attacks. Because the performance of really low power processors dramatically improves, future aggregator nodes will manage to performing modern-day data aggregation algorithms, thus making WSN less vulnerable. Thus, ascertaining standing of data and status of sensor nodes is vital for WSN. Such algorithms concurrently aggregate data from multiple sources and supply trust assessment of those sources, usually in a kind of corresponding weight factors allotted to data supplied by each source. To deal with this security issue, we advise a noticeable difference for iterative filtering techniques by supplying a preliminary approximation for such algorithms causing them to be not just collusion robust, but additionally better and faster converging.