Multi-source homogeneous continuous context inconsistency elimination algorithm based on improved basic probability assignment

In the dynamic and open environment, context-aware systems (CASs) obtain context information from dynamic, distributed and heterogeneous sources, but this context information usually has the inconsistency which would lead to inappropriate services. In this paper, a multi-source homogeneous continuous context inconsistency elimination algorithm based on the improved basic probability assignment (BPA) is proposed, aiming to solve the context inconsistency from multi-source homogenous sensors. In terms of the multi-source homogeneous context inconsistency elimination, we adopt sensor precision, membership degree and current data distance to modify the BPA on the basis of fully consideration of the frequency and amplitude of context. In the aspect of frequency, the precision is used to determine the probability. For the amplitude of context, we use the concept of membership degree from fuzzy set theory in the horizontal dimension and distances between the data at the current moment in the longitudinal dimension to assign the probability. The experiment results show that the continuous context inconsistency elimination algorithm proposed in this paper can significantly improve the accuracy of the context inconsistency elimination.