Lefebure.com / Articles / GPS - Past, Present, Future


The Beginning:
Back in 1957, the Soviet Union launched the first man-made satellite into space, commonly known as “Sputnik 1”. On the Sputnik satellite were sensors to monitor things in the atmosphere such as density and temperature. That data was being sent back to earth via a radio signal. After the Soviets announced to the world that they had put a satellite in orbit, electronics geeks around the world tuned their radios so that they could listen to the satellite. While most people were interested in the data from the satellite, a few people noticed that the radio frequency changed based on where the satellite was relative to their location on earth. This is known as the “Doppler Effect”. If you’ve ever stood along a road, you may have noticed that passing vehicles sound different when they are coming towards you than when they are going away from you. That is the same concept. It didn’t take long for scientists to figure out that they could measure where the Sputnik satellite was based on the frequency of the radio signal they picked up from it. That sparked an idea - instead of determining where the satellite was relative to a known location on earth, you could determine a location on earth from the known location of satellites. The concept of satellite positioning was born.

The first satellite based positioning system was functional in 1960. It was used by the US Navy. Several generations of systems have been launched since then, each with various improvements. The most significant improvement was the addition of an atomic clock on each satellite. 24 satellites are required in a constellation to be considered fully populated, which means that there will be a sufficient number of satellites visible anywhere on earth at any time. The GPS constellation reached full status in 1993, and currently has 31 healthy satellites in operation. The GPS system, as we know it today, is made up of satellites launched since the mid 1990s. The GPS system is owned and operated by the US Government’s Department of Defense.

The initial purpose of the GPS system was for military use. Other governments around the world quickly realized how useful this tool would be in the event of war. The Soviet Union started work on a global positioning system in 1976, and began launching satellites in 1982. In the event of war with the US, they wanted similar positioning capabilities. The name of their system, translated to English and abbreviated, is GLONASS. The Soviet Union’s system was completed in 1995, but then was abandoned when their economy collapsed. In 2003, Russia began restoring the system to operational status, and it currently has 100% coverage over Russia. A few additional satellites are needed for complete global coverage, which is expected to be complete in early 2011.


Recent Activity:
As of the writing of this article (March 2011), the most recent GPS satellite launch was March 24th, 2009. That satellite was the first to broadcast on the new L5 frequency band, but isn’t performing as expected. It has been in testing for two years now, and the next satellite launch is being delayed until the issues are resolved. The next launch is tentatively scheduled for July 14, 2011, but further delays would not be surprising.

Meanwhile, Russia has been aggressively building their GLONASS constellation. While the US typically launches one satellite at a time, Russia usually launches three at a time. The GLONASS launch on December 5, 2010 had a catastrophic issue with the rocket, and all three satellites on-board crashed into the Pacific ocean. There was another launch scheduled for late December 2010 that was delayed until February 2011, and it completed successfully. Russia is planning several more launches for 2011 including a new generation of satellite to replace some of the older ones.

Future:
The European Union is currently building a global positioning system known as GALILEO, and the first two satellites are already in space for testing. It will be at least three years before there are enough satellites in that constellation to be usable. Contracts have been signed for their ground based monitoring facilities and for additional satellites to be built and launched.

China is also building a global positioning system, known as COMPASS. Several satellites are currently in orbit for testing, but it is also several years away from being usable.

Ten years from now, we will have drastically more satellites available to use, which means better accuracy and fewer dropouts when working around trees and buildings. Today you can see, at best, 21 satellites at a time. Ten years from now, there will likely be 50+ satellites visible from your location, each broadcasting on at least two frequencies. In order to take advantage of the new signals, you can also expect to see new products from GNSS receiver manufacturers that are capable of processing 100+ signals simultaneously.


How global positioning works:
Imagine yourself at a sports stadium. There are speakers placed around the stadium so that everyone can hear the announcer. Every time the announcer says something, that sound comes out of all of the speakers at the exact same time. However, due to the speed of sound, you hear the sound from the speakers with various levels of delay. It takes longer for the sound to reach you from speakers that are far away. If you were to measure the time delay between the sounds from various speakers, you would be able to determine how much farther you are from those speakers. GPS works much like this, except that it uses radio waves instead of sound. Your GPS receiver “listens” for messages sent from the GPS satellites, and then measures the time delay between the various messages. The time delay is directly related to the distance. The messages that the GPS satellites send also contains information on where all the satellites are at. So if the location of every satellite is known, and your receiver knows it is some distance further from one satellite than another, it is able to determine where it is. I’m using two satellites as a simplified example, but in reality it takes a minimum of 4 visible satellites for a receiver to determine its location. To get decent accuracy the minimum is 5 visible satellites, and even that can be sketchy.

Why accuracy is difficult:
First of all, radio waves travel at the same speed as light, 186,000 miles per second, so this requires some very accurate time measurements. If you want a receiver that can be accurate to less than 1 inch, the receiver needs to be able to measure time delays of a tiny fraction of a second. At 186,000 miles per second, how much time does it take a radio wave to travel one inch? Answer: 0.000 000 000 08 seconds.

The second major issue is that the earth’s atmosphere distorts the radio signals. Radio signals will refract as they come through the atmosphere, much like light refracts through water or glass. To make this even more challenging, the amount of signal refraction in the atmosphere changes all the time. This is why DGPS and RTK systems are popular - they use one receiver that is stationary to measure the amount of atmospheric delay, send that delay info to one or more rover receivers, which then use that to calculate out the effects of the atmosphere.

Different levels of accuracy:
You can buy a GPS receiver for less than $50 today, but what do you get for that price? Accuracy within 100 feet. If you’re in an unfamiliar town, this level of accuracy is sufficient for providing directions to your destination. Most in-car navigation devices use a GPS engine that is specifically designed to work in environments where there is a poor view of the sky. Tall buildings, trees, and even the metal roof of the car can block satellites from view of the antenna stuck to your windshield. These receivers usually operate autonomously, meaning that they don’t use correction data from any base station. While this level of accuracy is fine for navigation in a car, it isn’t sufficient for agriculture.

Autonomous mode: South America is the single biggest market not covered by a free differential correction system. A good receiver running autonomously can get to within 5 feet most of the time. There have been some recent technology advancements in this realm. Different receiver manufacturers have different ways of accomplishing this, but it usually involves either averaging the numbers to “smooth” the data, or building models of the atmospheric delays. This data smoothing helps to reduce the short term position drift such as you would see between passes in a field. It doesn’t help with long term repeatability because there isn’t a stationary receiver to use as a reference point.

Differential Corrections: All differential correction systems use one (or more) base station(s) to monitor position drift. That data is then sent to the rover receivers to use. The data is typically sent to the rovers in one of two ways: via ground based radio towers, or via satellites. The US Coast Guard Beacon is a ground based radio signal for differential correction. When the data is sent via satellite, it is referred to as a SBAS (Satellite Based Augmentation System). The most commonly used SBAS in North America is WAAS (Wide Area Augmentation System).

What is WAAS? Like GPS, WAAS is a US government funded project. However, while GPS is a military project, WAAS was designed for aviation. In 1994, the FAA (Federal Aviation Administration) started a project to provide aircraft with increased accuracy and position assurance of GPS. The intent was to allow GPS to guide an airplane both in flight and during the approach for landing. An autonomous GPS receiver can potentially be several thousand feet off, which is not sufficient for landing an aircraft in the fog. The WAAS specification requires accuracy within 25 feet, which is reasonable for keeping an airplane above the mountain tops.

The WAAS system uses about 40 base stations spread around North America, all of which send data to three data centers. The data from all the base stations is combined to create a single stream of correction data that is useful over the entire continent. Those data streams are then sent to two satellite uplink sites, which send the data to three satellites. There is a lot of redundancy built in to the system. Europe has a similar system known as EGNOS, which is also designed for use by aircraft.

WAAS has been quite attractive to the agriculture market because it is free to use and is “good enough” for many of the tasks we do. Keep in mind that the WAAS system was built for aircraft, and the level of accuracy it provides is quite acceptable to that market. A high quality receiver using WAAS correction data will commonly have less than 8 inches of position drift over 15 minutes. Over 24 hours, the position drift will usually stay within a circle with a 5 foot diameter.

While WAAS is a single frequency (L1 only) correction system, other companies have built similar systems using dual frequency (L1+L2) base stations and more advanced correction algorithms. This results in better accuracy, both pass-to-pass and long term. The two companies doing this are OmniStar and NavCom.

OmniStar provides a variety of correction types. VBS is an L1-only correction similar to WAAS, but covers the globe, not just North America. OmniStar XP and HP are dual-frequency (L1+L2) corrections for the GPS constellation. OmniStar’s new offering is XP G2, which also provides correction data for both the GPS and Glonass constellations. XP G2 produces similar accuracy to XP, but the additional satellites keep you working in areas with a less-than-ideal view of the sky. Convergence time is also reduced by approximately 30% on XP G2.

NavCom is a division of John Deere, and they operate the StarFire correction system. SF1 and SF2 are both dual-frequency (L1+L2) correction signals available world wide. They were GPS-only since their beginning, but as of spring 2011, Glonass correction data is being added to both signals.

RTK: If you need accuracy down to the 1 inch level, then you need RTK. There are two main factors that are required to produce accuracy to this level. First, you need a reference station that is close to where you are working. Accuracy degrades when you get further from the base station because the atmospheric conditions can be different in different locations. Secondly, both the base and rover receivers have to be able to monitor not only the data coming from the satellites, but also the carrier radio waves that the data comes in on. For example, the L2 frequency is the encrypted military frequency band. Our civilian receivers can’t read the data on that channel, but we can monitor the raw radio waves. It is extremely difficult to measure the radio waves at this level, which is why RTK is the most expensive correction system available.

RTK+GLONASS: First of all, a GPS-only receiver will be able to see somewhere between 6 to 12 satellites at any given time. The satellites are moving, so that is why the visibility number changes through out the day. High accuracy position data requires a minimum of 5 satellites. If you are working during the part of the day when there are only 6 GPS satellites above the horizon, then you will be fine as long as you have a good view of the horizon in all directions. However, if you want to work near a building or trees, some of the satellites may be blocked from your view as they are hidden behind some object. Once you drop below 5 visible satellites, your receiver will not be able to report position data with high accuracy.

A receiver that can see both GPS and Glonass satellites will have approximately twice as many satellites available as a GPS-only receiver. While trees, buildings, and other obstacles can still block your view of some satellites, the idea is that there are enough satellites still visible for the receiver to maintain high accuracy.

Adding Glonass to a receiver does NOT increase the accuracy of the receiver. It only helps with availability due to the additional satellite visibility. If you are on the flat plains such as the Red River Valley, you probably won’t see any benefit from Glonass because trees and hills aren’t an issue. However, if you’re working anywhere near trees, hills, and buildings, Glonass will keep you on RTK when a GPS-only system can’t. For some people, lost productivity can be quite expensive. Glonass can reduce or eliminate the downtime while you wait for more GPS satellites to come in to view.




Sputnik Stats:
Launch Date: October 4, 1957.
Mass: 184 lbs.
Elevation: 134-583 miles.
Orbital period: 96 minutes.
Speed: 18,000 MPH

GPS Satellite Stats:
Mass: 4400 lbs.
Elevation: 12,550 miles.
Orbital period: 11h 58m.
Speed: 7,000 MPH

GLONASS Satellite Stats:
Mass: 1600-3200 lbs.
Elevation: 11,868 miles.
Orbital period: 11h 15m

Galileo Satellite Stats:
First two satellites launched October 21, 2011 from the French Guiana. Next launch: Summer 2012.
Mass: 1488 lbs.
Elevation: 14,429 miles.

WAAS Satellite Stats:
Mass: 4,000-13,000 lbs.
Elevation: 22,236 miles.
Orbital period: exactly 24 hours (geostationary orbit)
Speed: 6876 MPH
Trivia note: The FAA does not own these satellites. They are commercially owned and operated. The FAA just leases transponder space on them. They each carry numerous other signals.

Bizarre Trivia: Time is not a constant - the speed of time changes with both the speed of an object and the gravitational field around it. Each of the GPS satellites have four atomic clocks onboard. In order for the satellite’s clocks to be synchronized with clocks on earth, the satellite’s clocks have to be intentionally slowed during assembly on earth. Once launched into orbit, the clocks on the satellite appear to us as being the same speed as atomic clocks on earth. The difference is just 38 microseconds per day, but that is enough that it needed to be accounted for.


Last updated: October 30, 2011