Abstract: The location service aims to provide accurate real-time position information of target objects, which is based on positioning. However, the current underground positioning system has some problems, such as low positioning precision, poor real-time performance, insufficient capacity, limited database carrying capacity, only supporting one-dimensional positioning and so on. In order to avoid the influence of underground positioning system on location service, an underground location service system is designed. The system adopts a Docker-based micro-service architecture, which overcomes the problems of development iteration and performance bottleneck of the traditional monolithic architecture and looses the coupling between businesses. The system adopts the simultaneous ranging method of multi-label and multi-anchor nodes, which improves the ranging efficiency and the capacity of the positioning system while ensuring the ranging accuracy. The system uses multi-source data fusion positioning algorithm to improve the discrimination accuracy of the direction of the sign card relative to the anchor base station. The system adopts the positioning algorithm based on Kalman filter and weighted LM method and the low-complexity characteristic extraction method to optimize the positioning results, reduce noise interference, remove redundant data and improve positioning precision. The system introduces the time series database for mixed data storage, and stores time series data such as historical track in InfluxDB to improve system data access performance. The system adopts the publish-subscribe mode for asynchronous message transmission, which increases the reusability and sharing of public information. The system adopts Bearer verification for the location service interface to protect system data security and underground sensitive data. The practical application results show that the system can provide high-precision real-time position information of various underground targets, and provide important data support for the working face limit monitoring system, human-machine approach protection device, auxiliary transportation system and automatic driving system.