Evidential reasoning rule with dynamic reliability for performance assessment of wireless sensor network

Abstract

The performance of wireless sensor network (WSN) continues to change over time, and it is required to track the WSN performance accurately. In this paper, a new assessment model known as the evidential reasoning rule with dynamic reliability (ERr-DR) is proposed. In the ERr-DR, the evidence reliability evolves from the traditional static value to a dynamic value, which can reveal the impact of external noise on network performance reasonably. As an effective extension of traditional ER rule, the ERr-DR has excellent modeling ability. Furthermore, by analyzing the physical meanings of model parameters in detail, an optimization objective with targeted constraints is established. Compared with the existing optimization strategies, this not only improves the output accuracy of ERr-DR, but also maintains the interpretability of assessment results. Finally, A case study of a simulated WSN scenario validates the effectiveness of the proposed model.