Abstract:The researches show that there are various questions in the coal mine, such as the long-narrow underground roadway, significant the signal multipath effect, poor communication conditions, and so on. Underground wireless sensor is sensitive to the environment which will lead large error to the node location. On the basis of traditional RSSI algorithm, this paper proposed a trilateral weighted centroid location algorithm with a given attenuation calibration factor. Firstly, the measured attenuation value of RSSI is compensated to approximate the actual between the nodes. Secondly, the weighted centroid coordinates of estimates are calculated to be node coordinates after obtaining the estimates of unknown nodes by taking any three of the calibration distance. Simulation results demonstrate that, in the same experimental environment, the proposed RSSI algorithm performs smaller location error than the traditional, improves the adaptability to environment and positioning accuracy, and provides the basis for underground wireless sensor network node location.