Performance Analysis of 5G Positioning Procedures on Resource-Constrained Devices

Abstract

The fast positioning of targets in indoor environments poses a significant challenge on resource-constrained devices. Therefore, this article provides novel insights into fast positioning based on the fifth generation (5G) of mobile communication. Initially, a detailed introduction to 5G positioning focusing on a network-based approach that enables Enhanced Cell ID (E-CID) Location Information Transfer is given. To highlight this positioning approach, this article proposes an implementation of corresponding positioning procedures that are based on the New Radio Positioning Protocol A (NRPPa). Additionally, the performance of these positioning procedures is also investigated in terms of their time behavior using statistical methods. More specifically, based on a 5G positioning system, which is executed on two Raspberry Pi 5 single board computers, the Round Trip Time as well as the Periodical Time Interval metrics are examined depending on different parameters, such as the task priority, the thread pool size and the number of parallel executed positioning procedures. The results show that a resource-constrained Raspberry Pi 5 in combination with the proposed implementation is capable to handle the maximum of 256 parallel positioning procedures. In this case, however, the Round Trip Time increases by a factor of up to 829 compared to sequential processing. Furthermore, it is generally possible to transmit measurement data periodically, if the smallest Periodical Time Interval of 120 milliseconds is selected according to the 5G standard. In contrast to the Round Trip Time, the deviation of the Periodical Time Interval from the expected value can be kept largely constant independently of the degree of parallelism, if the thread pool size is decreased.