Qualcomm : A Q&A with Qualcomm Technologies inventor Stephen Edge on the development of location services and the future of 5G

Welcome to Qualcomm Invents. In this series, we're talking with some of Qualcomm's most prolific inventors about their work, the impact of their inventions, and their inspiration.

Location technologies are essential services in our modern lives. If you want food delivered via an app, need directions while driving, or require emergency medical care, your phone, tablet, or computer needs to access your precise location. There are also other use cases for location services tech that you may not be aware of. A retail chain, for example, may ask for your location during an online transaction to help identify a brick-and-mortar store nearest your home for curbside pickup. Location technologies are ubiquitous today, although they might not have been without Qualcomm Technologies' Principal Engineer, Stephen Edge.

Originally from England, Stephen attended Cambridge University, where he studied math and electrical engineering. He later earned a PhD in Computer Science from University College London, where he specialized in packet switching networks. He also ran simulations of the early TCP/IP protocol for ARPANET - the predecessor of the modern internet. Since joining Qualcomm in 2005, Stephen's work has focused on inventing the technological foundations of location services, as well as coordinating standardization efforts for 5G satellite communications.

We recently caught up with Stephen and talked about his groundbreaking work in the development of location services, the importance of satellite communication, unlocking the full potential of 5G, and his journey as an inventor.

The following conversation has been lightly edited for clarity and length.

Thanks for speaking with us, Stephen. This first question is pretty broad: Did you always know you wanted to invent things?

This may be somewhat unusual, but I didn't really appreciate what an inventor was until around 20 years ago. I never expected to be inventing or anything like that. When wireless technology was still evolving, I used to think being an inventor was something extremely significant and difficult. The things I was doing at that time did qualify as being inventive, but I wasn't aware of the true nature of invention such as the process of writing and filing patent applications. I think, in other companies, there are certainly many engineers that are not aware of their potential as inventors. Nowadays, I've been inventing, and I've had a large exposure to multiple inventions. I see now that inventing covers a much broader spectrum than people are aware of. I believe that many people who are not inventing things probably could, and they may very well have already created inventions but haven't recognized it.

Do you remember the first thing you invented at Qualcomm?

The first invention I worked on at Qualcomm was being able to periodically locate a UE ("user equipment"), such as a phone. Locating a UE is especially useful if you want to track a package or person, with a location being provided and updated frequently. The main issue was figuring out how to do it efficiently. How could we do that by efficiently reusing procedures that already exist to get just one location? When this question came up more than 16 years ago, there was nothing in the standards to define how to answer it. So, as I recollect, the first invention was coming up with new procedures to periodically locate a UE, terminal, or other device in an efficient manner and with low latency and high accuracy.

Would you tell us how your early work on packet switching networks and TCP/IP protocols informed your career?

No one back in the '70s or '80s expected the internet we have today to emerge - it was impossible to imagine. Personal computers barely existed back then, and nobody could predict that almost every person on the planet would either own or have access to some sort of computing mechanism. What was expected was that companies and organizations would be the main endpoints. There was something known as a Transnet that would be considered the precursor to the internet we now have.

The focus was the TCP/IP protocol. It was being deployed, but the deployment was slow because of the necessary re-engineering of the various hosts and nodes in the ARPANET. There were portions of it that weren't well understood. It wasn't clear how they should be used in an efficient and optimal manner. What I did was run simulations and stochastic models that would show how a TCP/IP connection between two endpoints would behave. The protocol uses a window, which basically defines the amount of data that one side can send to another before it can stop transmission. There was also a retransmission capability that was very significant.

Then there were aspects concerning transmission over multiple networks. Although the ARPANET was the backbone network, there were other networks attached to it, and there was a high value attached to enabling a TCP/IP connection to go across multiple networks. The internet we have now, in many ways, is based on what existed back then, but nowadays runs at a much higher bandwidth. Back then, the main links that connected different nodes and networks ran at speeds of around 56 kbps, which today is an amazingly slow speed even for one user. Back then, the limited bandwidth had to be used efficiently.

How would you explain the technology behind location services?

The basic idea is what you'd expect. When you're using your phone, there will be multiple times when you might need to know your position on a map or where to locate something nearby. You may be driving and want to know where the nearest gas station or restaurant is. Your location would be a prerequisite for providing you that information.

Now say you're looking to buy something. Well, it would be nice to know if there are stores nearby that have the products you're looking to buy. Again, that involves knowledge of your location. It comes up again and again in many ways. The important thing is that in each different case, your location can be obtained, and it will have a certain accuracy. Preferably, you'll want to have control over which entities are allowed to access your location. It's important that not everybody is always able to locate you. It eventually becomes a background capability that's used in many different applications. Sometimes it's visible, like when you see yourself on a map. Other times, it can be totally invisible.

What are the most significant real-world implications of these technologies?

I would say emergency services are the most critical. If you're involved in an accident, you expect medical services to arrive to assist you. At this point, providing a location to an emergency service is the most important thing. In any case, the location must be highly accurate. There's a requirement from the Federal Communications Commission (FCC) that you can always locate a caller (or their phone) within 50 meters accuracy for emergency purposes, and oftentimes you want to get even more accurate than that.

Another more likely scenario would be for navigation purposes. If you're in a car and need to be guided somewhere, your location will need to be highly accurate. You don't want to receive wrong directions. Location services are widely used now, and if they weren't available, there would be serious implications and a lot of services just wouldn't work properly.

How did your work in emergency services come about?

Back around 1998 and 1999, there was a requirement from the FCC that anyone who dialed 911 could be located and that location could be sent to public safety services so they could transmit it to emergency services. It could be for the police, a fire department, an ambulance, or even a mountain rescue service - there were a lot of different possibilities. The Alliance for Telecommunications Industry Solutions standards body that had already existed in the U.S. decided to support this. There was about a two-year period where we had to invent the capabilities and standards to support this initiative.

I was very heavily involved in this and helped chair a group that brought in multiple contributions. This is how I started out in location and started inventing. I didn't realize then that it qualified as an invention, so I didn't file any patents initially, and a lot of things that went into standards ended up as free for everyone to use. I thought the work was both publicly and socially useful, so it was worth spending a lot of time on it. Back then, we all thought location services would basically disappear as something that would need to be researched and enhanced, once the emergency services support was complete. We were all wrong, of course, as it has continued to evolve since then, mainly due to its use in commercial services and continued improvement for emergency services such as when a 911 caller is indoors.

Tell us about your work coordinating standardization for 5G satellite communication.

One thing about wireless networks such as 4G and 5G that perhaps many people can relate to is that the coverage is not at all continuous or ubiquitous. In certain locations, oftentimes you can have difficulty getting coverage. For instance, if you're staying at a hotel or in your home and move from one room to another, you may have a drop in connectivity. That's an issue that may never disappear. The reasoning behind this is that operators can't afford to install base stations in every possible location, it just isn't feasible.

Additionally, wireless signals can be blocked by different obstacles and there are many areas (e.g. in remote locations and less developed countries) where coverage is spotty or not available. The benefit of satellite communication is that it's something that can provide coverage virtually everywhere - outdoors and to a degree, indoors. Almost anywhere outside in the world, with satellite communication, you could have coverage. Particularly for countries with less developed infrastructure, it could be exceedingly useful. After enabling 5G, instead of base stations, you could have the network connected to low earth orbiting satellites. Nearly anywhere you could think of, you could be in the coverage area of at least one satellite, if not a few of them. In five to ten years, you could be almost anywhere on the Earth and have wireless communication capability. For a lot of people, it could make a huge difference.

5G is predicted to enable up to $13 trillion in global economic activity by 2035. What do you see as the key to unlocking the full potential of 5G, and how does your work intersect with that?

5G is already on a major trajectory. There are many different organizations with a strong vested interest in it. It's already well underway. In concrete terms, the operators will be building out their networks. They'll install 5G base stations in some cases by just converting 4G or 3G base stations into 5G. One of the critical things about 5G is to increase the capacity to enable higher bandwidth and to reduce latency, allowing you to send more information faster. More base stations will help allow a higher network capacity, which will help allow for better coverage. The result will include 5G becoming more apparent over the next couple of years, with a much higher capacity. It will help enable better coverage than existing 4G networks, and you will be able to do more in terms of being able to transfer more data with shorter delays.

What's the most fulfilling part of your job?

What I like is finding a good solution to a difficult and important problem. It's not something that happens every day. It's maybe something that happens every couple of months. Finding important solutions is a process. First, you need to come up with a basic idea for the solution, then develop that idea and identify and overcome any limitations to enhancing it and extending it into a full solution. Next, you may need to introduce the solution into standards. While you may have a good solution, you then must convince other people that it's worth supporting, as there may be other competing solutions. If you have the best solution, you can often succeed in getting your solution agreed upon if you are persistent enough.

What's the most exciting or consequential discovery you've made in your career thus far?

One I think that's the most consequential may not even sound like much of an invention, but it certainly is having a lot of impact. Around 20 years ago, location services were supported using protocols. The protocols are used between the UE, that's a terminal, and a location server. When we started, it was basically GSM (2G). We defined these protocols and methods of locating the device. Then we did something similar with our work on 3G. The idea that I came up with was to have a protocol that could be used for all iterations of G, so it could be applied to 4G, 5G, and in the future, 6G. It could support positioning techniques or location techniques. My idea was that you could combine it all into a common generic protocol.

The benefit of this idea is that you get a reusable capability. Instead of having to define and implement something totally new, you can take something existing and just extend it. It makes processes much more efficient and flexible, as it allows terminals and location servers in the network to be able to select things which they both support and not have to switch protocols. They can just use one protocol. That idea ended up being patented. It's something that should remain applicable to future development. There are probably more than a billion terminals now that are supporting this.

What are you most excited to work on in the future?

I tend to take things as they come. I see areas that I think are worth improving so we end up with better services and capabilities. Since I work in standards, we standardize things that will be deployed in several more years. Right now, the 5G satellite that I discussed, you'll probably not see for another couple of years but may thereafter be common (just like Wi-Fi) on nearly all new phones. In terms of standards, we must deal with it now. When I look ahead to the future, I'm looking a year or two into the future to standardize things that will be deployed in three or four years.

How has Qualcomm supported you in your efforts? How would you describe the culture of innovation at Qualcomm?

Qualcomm encourages people to invent things. It supports inventors in the sense of giving them time to work on their inventions. Qualcomm acknowledges that besides coming up with an invention, you must spend time writing it out properly so others can understand it, filing patents, and introducing it into standards. Qualcomm knows that this is needed and supports its inventors in doing that.

What advice would you give to other inventors or aspiring inventors?

I would give two pieces of advice. First: Anyone can invent something. Not everyone is equally good at inventing, but any person can invent at least one or several things. What's important is finding the best way to invent and to start doing it. Don't think about inventing, just start doing it. Inventors learn as they go along, and the more they invent, the more they learn. Second: Associate yourself with experienced people, those who are willing to assist you in your endeavors. If you examine the process of someone who's experienced, it'll make your inventions easier to develop.

Learn more about the culture of innovation at Qualcomm:

Read more about other inventors at Qualcomm:

Ananth Kandhadai invented technologies in signal processing, speech compression, and computer vision.

Lin Lu pioneered the invention of Smart Transmit - a breakthrough technology that optimizes uplink speed and coverage for 5G.

Dr. Simone Merlin helped make Wi-Fi faster and more efficient for high-traffic networks.

Dr. Lola Awoniyi-Oteri helped optimize 5G connectivity for power efficiency and mobility.

Dr. Tingfang Ji helped make it possible for 5G to scale across different industries and fields.

Dr. Marta Karczewicz helped make it possible to stream, share, and communicate over video.