There are a lot of potential fault information in the vibration signals of industrial equipment, such as CNC machine tools and construction machineries. Under the background of industry 4.0, it is necessary to send the vibration signal of equipment to the Internet of Things( IoT) to realize the remote online monitoring function. Since the equipment vibration signal is continuous time domain signal, if the data generated by the vibration sensor was directly uploaded to the cloud server, the following problems will occur at least: 1) a large number of unprocessed data will occupy a large amount of storage space of the cloud server. 2) The data flow is too large and the data upload cost is too high. In view of the above problems, the Micro-Electronic Mechanical System ( MEMS) vibration accelerometer was used to pick up the vibration information of the equipment . The high-speed 32-bit Micro Contro Unit (MCU) named STM32 was employeed to carry out Fast Fourier Transform (FFT) on the vibration signal, and only three feature parameters were uploaded to the IoT platform named OneNET through CAT1.0 mobile network: the main frequency of vibration , damping ratio and vibration intensity. A wireless vibration monitoring experiment of a cantilever beam is carried out finnally, and the theoretical analysis is consistent with the experimental results. The problem of blocking and pressure caused by excessive vibration data to the IoT was solved well.