Algorithm of Context Inconsistency Elimination Based on Feedback Windowingand Evidence Theory for Smart Home

Due to the accuracy difference of sensors, equipment heterogeneity, network delay and diversity of algorithms etc., inconsistency of context information may emerge, which would affect the accuracy and reliability of context reason- ing and decision making. To solve this problem, an algorithm of context inconsistency elimination which is based on feedback windowing and evidence theory is proposed in this paper and can be applied in many scenarios, like smart home. Simulation results show that the proposed algorithm outperforms the other counterparts in accuracy when errors especially burst errors occur, and it can improve the accuracy and reliability of context-aware system.