From 24-28 September was the ION GNSS+ meeting held in Miami, Florida. Evolutions in the conference name (now including the “+”) are well reflected in the content. Research on precise point positioning (PPP) is now driven by the automotive industry which calls for less stringent accuracy needs but higher integrity. This blog post summarizes what I consider to be the key points of the conference from a PPP perspective, with an emphasis on low-cost devices and integrity.
Raw GNSS measurements, available from Android devices since 2016, have sparked an interest in many researchers. Two complete sessions were dedicated to this topic in the meeting which, as most people seem to agree, might have been too much. There were countless presentations showing noise figures of carrier-phase and pseudorange measurements collected with many smartphone/tablet models. As we know, smartphone antennas are linearly polarized and contribute significant multipath errors. Authors excluded multipath using several approaches (special box with re-radiated signals [Geng et al. 2018], anechoic chamber [Roberts et al. 2018], etc.) to truly assess the noise characteristics of the GNSS chips, which is still at the meter level. The release of Android P on August 6 2018, having the option to deactivate duty cycling, occurred too late in the year to allow most authors to fully exploit this feature.
The key point regarding low-cost devices this year was the release of the Broadcom BCM4775 chip, providing dual-frequency code and phase measurements (L1/E1 and L5/E5a). Since this chip is already included in the Xiaomi Mi8 phone, initial results with this device were presented by a few researchers. While the antenna is still corrupting measurements with multipath, it was reported that the noise contribution from the chip itself was around 2.5 m for L1/E1 and 1.5 m for the L5/E5a signals. This performance is a clear improvement over the L1-only chips currently on the market. As a result, centimeter-level horizontal accuracy using short-baseline RTK could be attained in minutes [Warnant et al. 2018]. This is clearly a step forward and will fuel a lot more research for next year…
PPP for the automotive industry
The second topic of interest was integrity for PPP. Trimble added integrity checks to their RTX service by checking for errors in their orbit/clock, tropospheric and ionospheric corrections largely exceeding their expected precision [Weinbach et al. 2018]. This is achieved in two stages, through pre- and post-dissemination tests, to monitor the whole chain of corrections, from generation to distribution. Interestingly, no integrity alerts are sent to users for errors smaller than 50 cm, which seems to suggest that integrity is focused on automotive applications rather than high-accuracy users.
Sapcorda (a consortium including u-blox, Geo++, Mitsubishi and Bosch) gave an overview of their proposed service, emphasizing how integrity and certification is their added value to the GNSS correction market [Leandro et al. 2018]. Since industry seems to be delaying the efforts of the RTCM-SC104 in adopting new state-space representation (SSR) messages, Sapcorda also proposed a new format for the exchange of real-time GNSS corrections [Vana et al. 2018]. Although details on this format are still not released (the Sapcorda website has, as of today, no information on this topic), the paper announced that orbit/clock messages would allow for a 40-50% bandwidth reduction with respect to (uncompressed) SSR messages, and would support atmospheric (tropospheric and ionospheric) corrections. Integrity is also an announced feature of the format, although I haven’t yet fully grasped the concept, apart from the possibility of sending the same messages several times to maximize reception at the user end.
It is clear that NovAtel is also targeting the automotive industry as it demonstrated, in March 2018, testing of their GNSS positioning engine and correction services with GNSS chipsets provided by STMicroelectronics. In Miami, they announced the upcoming release of the Terrastar-X service providing regional ionospheric corrections for fast PPP convergence [Jokinen et al. 2018]. NovAtel also partnered with Stanford University to investigate integrity at the user end [Gunning et al. 2018]. The authors applied the RAIM concept of solution separation, consisting of determining the maximum distance between solutions computed by excluding satellites from the PPP filter. Even though the results presented seem to provide satisfactory results in several applications (static, car navigation, airplane), I could not fully understand from the presentation if a single faulty satellite was assumed or if multiple faults could be accommodated simultaneously. With the latter, and considering a multi-GNSS solution, the number of parallel PPP filters to run would grow rapidly, making the concept hardly applicable in practice. I will definitely look forward to reading the full paper if it becomes available this week.
Please keep in mind that this summary is based on my own interpretation of the presentations from the conference. I encourage everybody to read the conference proceedings for more details!
Geng J, Li G, Zeng R, Wen Q, Jiang E (2018) A Comprehensive Assessment of Raw Multi-GNSS Measurements from Mainstream Portable Smart Devices, ION GNSS+ 2018
Gunning K, Blanch J, Walter T, de Groot L, Norman L (2018) Design and Evaluation of Integrity Algorithms for PPP in Kinematic Applications, ION GNSS+ 2018
Jokinen A, Ellum C, Webster I, Shanmugam S, Sheridan K (2018) NovAtel CORRECT with Precise Point Positioning (PPP): Recent Developments, ION GNSS+ 2018
Leandro R, Graf zu Eulenburg B, Gonzalez P (2018) Safe GNSS Augmentation for High-precision Autonomous Systems - Design and Performance, ION GNSS+ 2018
Roberts W, Critchley-Marrows J, Fortunato M, Ivanovici M, Callewaert K, Tavares T, Arzel L, Pomies A (2018) FLAMINGO - Fulfilling Enhanced Location Accuracy in the Mass-market through Initial Galileo Services, ION GNSS+ 2018
Vana S, Aggrey J, Bisnath S, Leandro R, Gonzalez P (2018) Analysis of GNSS Correction Data Standards for the Automotive Market, ION GNSS+ 2018
Warnant R, Van de Vyvere L, Warnant Q (2018) Positioning with Single and Dual Frequency Smartphones Running Android 7 or Later, ION GNSS+ 2018
Weinbach U, Brandl M, Chen X, Li Y, Pastor F, Rodriguez-Solano C, Reussner N, Seeger S (2018) Integrity of the Trimble RTX GNSS Correction Service, ION GNSS+ 2018