Multiple level parallel decision fusion model with distributed sensors based on Dempster-Shafer evidence theory

The application of Dempster-Shafer evidence theory to decision fusion with multi-sensors is discussed in this paper, and whose analysis focuses on comparing and contrasting Dempster's rule of combination and its improved model. Considering such defects of the Dempster combination rule as failing to support the totally inconsistent evidences and being insensitive to the base of aggregate, and also due to the lower border of belief interval in the improved model of Dempster's rule being too low, we put forward a new multiple level parallel decision fusion model with distributed sensors based on Dempster-Shafer evidence theory, accordingly. The results of simulation experiments demonstrate that the model we present is more feasible and effective, and will be of great significance to solve the problem of military target identification.