Secure trust reputation with multi-criteria decision making for wireless sensor networks data aggregation

The presented approach is a step towards a more trustworthy sensor in-network data aggregation without the need of hardware tamper-protection or other high cost modifications on sensor node equipment. On the one hand data aggregation is often used to reduce energy consumption but on the other hand this often enables attackers to compromise the network by simply altering data or disrupting transmissions. Most secure data aggregation protocols can detect manipulation of the aggregation results only based on one criterion. In our approach multi-criteria decision making based on a weighted Choquet Integral is possible. Thus, continuous sensory data processed by our trust level calculation algorithm can be one criterion next to criteria like confidentiality, etc. or even data correlation of different kind of sensory to decide how trustworthy a sensor node is. The novel trust level calculation algorithm is based on a Gaussian probability function paired with a Byzantine decision making.