Fuzzy Analysis For Assessing Trust Space Within Wireless Sensor Networks

We developed a fuzzy logic framework to analyze the trust of an outdoor deployed data sensor (DS) within a wireless sensor network (WSN), where we selected three trust factors: weather index, battery residual and group deviation as input membership functions and generated trust values as an output membership function. A combination of internal (battery residual) and external (weather index) parameters along with group deviation from its neighbors are selected as trust factors, since the performance of outdoor deployed sensors are often affected by these uncertainties. A combination of trapezoidal and triangular membership functions with five linguistic variables and a defined set of rules are applied to Sugeno Fuzzy inference systems (FIS) to assess the trust. The fuzzy framework verifies the sensors trust at the gateway prior communicating the data to the central server, thus avoiding energy consumption on untrusted data. The simulation results revealed that Sugeno FIS demonstrated 36% energy savings against normal transmission by hopping untrusted data.