Multi-Scenario Dynamic Indoor Positioning Based on 5G Commercial DAS Signals

The fifth generation (5G) network is developing rapidly, and the user equipment (UE) around the world has increased dramatically. Currently, 5G signals can be effectively received in the most indoor scenarios. Thus, research on indoor positioning with 5G signals has its potential promise in terms of low cost and high usability. Primarily, the commercial deployment of 5G base station has two types, i.e., a traditional 5G Distributed Antenna Systems (DAS), and a newly updated system called 5G digital indoor distribution. Our previous work has investigated indoor positioning with 5G indoor digital system. In the meanwhile, it has been noticed that, 5G DAS still works well in many public indoor scenarios, such as office building, railway station etc. Therefore, in this work, we study the indoor positioning with 5G DAS commercial signals. Reference signal received power (RSRP) based on multi-beam channel characteristics is identified as fingerprint for positioning. During the indoor tests, dynamic fingerprint features are collected, and a motion-constrained fingerprint matching model is constructed by using the LSTM algorithm. Field indoor tests are carried out in different indoor scenarios to verify the effectiveness of the method. The test results show that the method is able to achieve the positioning accuracy with the error of 3.5 meters with the 68 % confidence.