A review of the possible sources of common GPS errors
Today, GPS applications and services are everywhere, and accessible to almost everyone.
In the UK you can now purchase a portable “SatNav” system, admittedly from a manufacturer you might not have heard of, for less than it costs to fill your car with fuel! This is great news indeed, right? Well, maybe … but then maybe not. When was the last time you opened your morning paper and did NOT find an article about a lorry blocking a country road or a car getting stranded somewhere remote? Many of us in this industry, are concerned about how this bad PR about “SatNav” might end us all up in a bad place.
Being in the GNSS industry, we know, of course, that the GNSS itself is not to blame for these misdemeanors. We know that the GPS receiver is reporting a position accurately on flawed route instructions. The GPS accurately takes the lorry to the country road. We also know that this issue is being actively worked on and, hopefully solved. Improvements underway include smart routing to avoid minor roads, systems that take the size and weight of vehicles into account, and improved mapping information. These issues are squarely in the commercial domains, where the customer demands perfect performance for extremely low investment. In more professional applications, it’s usually reasonable to assume that the user will want to understand the limitations of the technology. The golden rule is not to rely slavishly and without thought in the GNSS position. The position you see may not always be entirely accurate or without risk.
A comprehensive review of the possible sources of GPS error are well beyond the scope of this article. However, we can touch here on some of the most common, and not always most well known, error sources. Looking at an overall “error budget” for a single frequency GPS receiver in an open sky environment, the major component of error is atmospheric. The L-band GPS signals will be delayed dependent on the electron count of the atmosphere. GPS receivers attempt to back-out these atmospheric errors via standard models in the receiver itself. The most common is known as the Klobuchar model. These models are quite good at giving us sub five meter accuracy we expect today. In fact today the accuracy is often even better than this, say 2-3 meters, partly due to us currently being at in the middle of an 11-year solar cycle. However the high solar activity is more difficult to model and will cause higher inaccuracy in single frequency receivers. For dual-frequency receiver users the problem is much less, as the atmosphere affects the two frequencies differently and it is therefore possible for the receiver to measure, and back-out, the ambiguity of the atmosphere from the receiver’s position determination.
Multipath contributes a much more dynamic and potentially disruptive element to receiver errors. Multipath occurs when a signal arrives indirectly at a receiver antenna. Because of the indirect path, the signal length is longer. While receivers try to identify and reject multipath signals from their position calculation, this is not always successful. Multipath effects can introduce high magnitude errors, of the order of tens of meters not being uncommon. In addition, for moving vehicles, multipath can vary quickly, introducing see-sawing effects in speed and position. Bear in mind also that multipath will never be the same twice, even at the same location, due to the ever changing positions of the GPS satellites in the sky. Very occasionally, GPS has been known to have signal transmission errors. Correction messages, based on reference data from GPS monitoring stations, are often used to inform receivers of a bad satellite or signal condition.
There can be a lag between an error occurring and a health flag being set and transmitted to the receiver. The good news for all of us is that, used with a degree of knowledge, GPS is an efficient and reliable professional tool. With the addition of new GPS signals (e.g. L2C and L5) multi-frequency professional receivers will be much more readily available. New GLONASS satellites have been added quite recently. And Galileo and potentially the Chinese Compass signal will also be available during this decade We still have some way to go, particularly in relation to navigation anywhere, but exciting times indeed are ahead for those of us tracking GNSS technology developments.