Assessing the performance of raw measurement from different types of smartphones

Abstract

In recent years, the API level 24 allow the user have access to the GNSS raw measurements including pseudorange, doppler and carrier phase, which opens a new era for precise navigation on smart devices. As we known, many manufactures produce smartphones that access raw GNSS measurements, like Google, Huawei, Samsung et al. Different manufactures may use different GNSS chipset, which may result in different performance of measurements and navigation. The objective of this paper is to assess the performance of different types of smartphones including Huawei Mate9, Huawei P10, Samsung Note8, Google Pixel2 and HTC U11+ in open sky area. All of these smartphones are declared to be able to track GPS, BDS, Galileo and GlONASS satellites and access the pseudorange, and the first-three can also output the carrier phase measurements . Firstly, the tracked satellites number, SNR and their NMEA results were compared with that from nearby Trimble-R10. The result shows that each types of smartphone has its advantages and disadvantages. The Huawei Mate9 and Huawei P10 tracked fewer BDS and Galileo satellites, but they can track many satellites from GPS and GLONASS. For Google Pixel2 and HTC U11+, BDS satellites can be tracked only sometimes and also some Galileo satellites for Google Pixel2. Samsung Note8 can receive satellite data from all operational GNSS systems stably, though the satellites number is fewer for each system. For Samsung Note8, Google Pixel2 and HTC U11+, the SNR of most trackable satellites is above 20 and just few of them are between 10 and 20, while for Huawei Mate9 and Huawei P10, many satellite can be tracked whose SNR is between 10 and 20. Statistics results of the NMEA show that the accuracy of NMEA from Huawei Mate9 and Huawei P10 is much better than that from the other three phones. The accuracy of NMEA form Huawei P10 is highest and its RMS are 1.77m and 0.99m in north and east respectively. Secondly, The analysis of double-differenced pseudorange residual indicates that the UTC 11+ performs worst and its double-differenced residual varies within ±50m, the other four phones are almost at the same level and vary within ±30m. Double-differenced carrier phase residual shows obvious duty-cycle for all these smartphones, which limit the use of carrier phase. As different smartphone has different implementations of duty-cycle, the clock behaviors were analyzed and compared with each other, which allow us characterize the duty-cycle of these devices. We concluded that the higher precision compensated for temperature variations (TCXO)Hardware clock is not continuous (turning on and off) during the non-tracking periods for Google Pixel2 and HTC U11+, while for Huawei Mate9, Huawei P10 and Samsung Note8, the TCXO is turned off and low-power consumed crystal oscillator (XO) is providing time. After a comparison of the positioning result with raw pseudorange measurements, we did not find obvious difference for the five devices, then the doppler measurements are introduced to smooth the pseudorange and it can improve the accuracy effectively.