Expert Advice: Taking Up Positions — Galileo and E112

By Andy Proctor

Sessions on indoor navigation and a keynote from Google at February’s International Navigation Conference (INC15), organised by the Royal Institute of Navigation, addressed the revised E911 positioning requirements in the United States, and flowed over into speculation about E112 emergency calling parameters in Europe’s near future.

According to the 2014 U.S. Federal Communications Commission report, 75 percent of 911 calls now come from mobile phones, more than half of those originate indoors, and around 1 percent of emergency calls contain no location information from the caller (due to distress, confusion, language issues, illness, and so on). The report estimates 10,000 deaths per year in the United States might have been avoided if a landline had been used instead, since location information for landlines can be provided confidently.

Discussion in the breaks of INC highlighted a misunderstanding amongst some parties that E911 mandates the use of GPS for position location determination. In fact,  E911 does not mandate any specific technology; it specifies performance criteria in terms of accuracy that must be met. The recently revised performance criteria include indoor performance, and some of the technology discussed at the INC is able to meet these requirements without using GNSS at all.

This could be troublesome for Europe, which is looking at the imposition of Galileo as part of an A-GNSS technology push for the E112 application. The real problems, discussed during INC and in European consultation processes with safety of life services such as E112, are:

• the accuracy of the position derived by the device and/or network, and
• the timeliness of the delivery of that position to the Public Service Answering Point (PSAP).

The E911 directives address these points directly, and the infrastructure in the cellular networks is in place. Does simply implementing a Galileo capability into a European mobile device solve these problems?

In many outdoor cases, implementing Galileo can bring benefits, including signal diversity. And of course the E112 proposal is greater than just “adding Galileo.” It does address the second problem of timeliness of delivery and data transfer, but there are significant infrastructure upgrades required across Europe for the provision of this location data to the PSAPs.

What the E112 processes do not currently do is specify performance criteria for the position location accuracy. This means that the position estimate provided under E112 is likely to be a cell-ID fix, with an accuracy ranging from hundreds of meters to dozens of kilometers.

Galileo on Mobiles. Further discussion during the conference delved into the realms of the specifics of implementing A-GNSS, including Galileo, onto a mobile device. Conversations centered around if any future E911 or E112 positioning capability would be aligned around a single-chip solution as generally currently deployed on a device, or if some of the functions will be moved up the stack into the operating system (OS) of the device, into software.

Most opinions were against this latter concept, and a panel at the ION GNSS+ last year in Florida concluded the same thing. However, questions were asked about some ideas relating to identifying the emergency number at the time of dialing and then starting the position location determination functions in readiness for the need to provide the device location. This addresses the first bullet point earlier, the accuracy of the position derived by the device and/or network. If this is carried out in the OS or software layers, vulnerability of the system will be increased overall as the OS of a mobile device is a target for the cyber criminal community.

A robust software-based solution is, however, being rolled out in the United Kingdom in the form of eSMS, bringing mobile operators, government and handset vendors together to provide location data via SMS to the PSAP. The advantage of this approach is that no new standards or major infrastructure changes are required, and the time to implement is small.

Further discussions established that future chipsets are likely to use whatever GNSS signals are available, regardless of whether they are GPS, Galileo, GLONASS, Beidou and so on. This, coupled with new signal processing techniques (single-frequency observable for example), increasing sensor clustering on devices, and user demand for services, may make the use of a specific GNSS system above others somewhat redundant. Certainly picking up on a point made by Chandu Thota from Google, GNSS is “not relevant” for their indoor positioning solutions, and technologies they are working on, in both hardware and mapping improvements, are looking at meeting indoor accuracy requirements down to a target requirement of 1 meter, without GNSS.

Taking these points into account, questions were asked from the floor of the conference about the legal position of the EC mandating Galileo as a positioning method as well as the willingness of the global mobile chipset and device industry to be told what to do. Perhaps specifying strong performance criteria, as in the United States, is the way forward to “reboot” the European E112 system. No one disputes that a properly functioning E112 is a life saver and a good thing to do; however, the points discussed here detail some of the concerns expressed during and after hours at INC15.

In February 2015, the Royal Institute of Navigation hosted the International Navigation Conference in Manchester, UK. Keynotes at this well-attended conference included Harold Martin, director of the GPS Coordination Office; Gian Gherardo Calini, the head of market development at the European GNSS Agency; Todd Humphreys from the University of Texas; Chandu Thota from Google; and others. The conference covered multiple technology tracks including indoor navigation, autonomy, quantum technology and the resilience of GNSS systems.

Andy Proctor is lead technologist for satellite navigation at InnovateUK, the UK’s innovation agency. He acknowledges Ramsey Faragher, Cambridge University, for help in the preparation of this article.