In order to improve the fault diagnosis accuracy of rolling bearings that are commonly interfered by gear noise and variable rotating speed,a fault diagnosis method based on the ensemble empirical mode decomposition ( EEMD) and the instantaneous dominant meshing Multiply ( IDMM) is proposed.The proposed method employs EEMD to eliminate the gear noise in bearing vibration signals,transfers IDMM into an instantaneous rotating frequency,and uses the order tracking to overcome the decrease of diagnosis accuracy due to rotating speed variation.Simulated and experimental results show that the proposed method is effective in eliminating the interferences caused by gear noise and rotating speed variation,thus improving the characteristic order amplitude of faults.Accordingly,the installation of tachometers can be avoided.