Insights and FAQs

Satellite positioning (also known as GPS or GNSS) works by using timestamps broadcast by satellites.

GNSS receivers use timestamps from at least 4 satellites, combined with the known satellite position in space, to calculate the receiver’s position on earth.

This works perfectly when the satellite signal is unobstructed - in direct Line Of Sight (LOS).

However, most GNSS-enabled devices these days are used in more developed areas, where satellite signals can be reflected or obstructed by objects such as buildings, vehicles, landmasses and even vegetation. When receivers pick up a reflected or Non Line of Sight (NLOS) signal, the timing message will be older than it should be (echos always arrive late) and so the subsequent estimation of distance to the satellite will be too long, resulting in an inaccurate positioning fix. Even if the line of sight signal is available, any reflected copies can interfere with it when they are all picked up by the receiver together, corrupting the true measurement, making it look either early or late through distortion of the signal. This problem, known as multipath interference, is especially problematic in cities and towns.

Multipath interference and NLOS signals can result in positioning fixes in cities being wrong by over one hundred metres.

Supercorrelation™ is a software upgrade for the receiver and solves the problems of multipath and NLOS interference by determining the angles at which satellite signals should arrive at the receiver. By comparing the measured angles of arrival with the expected angle of arrival, Supercorrelation™ can flag and reject any signal not deemed to be true. The receiver can then focus on the LOS signals, even if they are very faint. This patented technology is the only software solution that is able to do this using the very low cost hardware used in consumer devices.

Correlation is the process by which receivers detect and measure signals from specific satellites.

In order to process satellite signals, GPS receivers try to match, or “correlate”, the incoming signal with a local copy of each satellite’s unique code. When the signal lines up correctly with the local code then there is a peak detected in the correlation process.

The arrival time of these correlation peaks measured by the receiver are used to calculate the user positions.

Multipath interference and the presence of reflected signals corrupt the shape and exact arrival time of these correlation peaks, resulting in an incorrect range estimate to the satellite.

The Supercorrelation™ process changes the local copy of the satellite’s unique code in the receiver during the correlation process to a much longer and more complicated version than has ever been used before in radio positioning. This allows the receiver to distinguish the subtle differences between signals arriving from different directions, but building very long correlation sequences is in practice difficult. The Supercorrelator must correctly compensate for any ongoing motion of the receiver at the centimetre level (think about just how complicated a pattern through space your wristwatch “draws out” in space over one second while you are out on a run), and must also account for variations in the stability of receiver components over time. Supercorrelation™ accounts for these user motions and the low quality components, and it is these adaptations to the local copy of the codeword replica that allow the resulting version to be sensitive to angle of arrival.

For a more technical explanation, view the attached webinar on Supercorrelation™ by FocalPoint Founder Dr Ramsey Faragher.

Supecorrelation™ technology removes sources of error such as multipath interference and non-line-of-sight signals. Its effect is most apparent in areas where these error sources are dominant (such as in dense cities) and position fixes can be 50-100m away from the true location. We have conducted a number of trials in various cities around the world with major GNSS chipset companies and demonstrated using their own hardware a significant reduction in the tails of the positioning error distributions. This RIN webinar demonstrates a trial around London for example where the 95% confidence bound in the position accuracy was improved by roughly a factor of 10.

The exact figure for trials depends on the density of the environment but is typically in the range of 7-15x improvement in urban canyons. Other metrics such as Doppler accuracy and positioning integrity are also improved by a similar degree. Sensitivity can be increased by 5db - 15dB depending on the receiver design.

The most extreme errors that can be created by coherent multipath interference on an L1 GPS C/A code signal is around ±150m. Supercorrelation™ can remove this error, leaving the residual fundamental error associated with accurately measuring the centrepoint of the correlation peak, which is around 3-5m. So the fundamental theoretical accuracy improvement for Supercorrelation™ in the most extreme example of multipath interference is around 30x.

Commercial implementation of Supercorrelation™ requires integration with the GNSS chipset. However, representative trials can be performed via post-processing using recorded satellite data. To perform trialling in a chosen location, please make an enquiry.

Supercorrelation™ enables the receiver to determine the angle of arrival of all radio signals. This means that a spoofer signal can be recognised due to the location from which it is broadcasting.

All Supercorrelation™ products reject spoofer signals as standard. Our Spoofer protection products include the ability to flag for spoofer attacks and to localise spoofer signals - enabling users to identify where the spoofer signal coming from.

Supercorrelation™ does not include a Navigation Engine - it is designed to be embedded at the deepest level of the receiver, and to output data that can be used by any navigation engine.

Skyscan is a patented technology and is an example of an entirely new class of GNSS measurements that Supercorrelation™ enables. It depicts a receiver’s complete view of its surroundings, showing the azimuth, elevation, and strength of all incoming signals a bit like a radar scan.

This data enables us to determine the angles of arrivals of the signals, and enables receivers to reject unwanted signals such as reflections or spoofers.

RTK or Real Time Kinematic improves the accuracy of GNSS receivers by broadcasting correction messages from a static base station located in an area with low multipath and NLOS problems.

While RTK depends on this additional hardware, the receiver’s proximity to it, and a guaranteed radio connection between the two, Supercorrelation is a software upgrade inside the receiver that requires no additional aiding or hardware.

GNSS receivers with Supercorrelation™ integrated within them can still use RTK, and Supercorrelation™ itself can improve the performance of RTK systems by ensuring that only line-of-sight signals are used in the RTK calculation (by filtering all incoming signals by angle of arrival).

3D Map aiding (3DMA) has been developed by various entities, including Google, as a method for solving the problem of poor GNSS positioning in cities. It uses additional 3D model data about buildings surrounding the receiver location, and attempts to use the knowledge of building locations to improve a receiver’s position estimate. For example if a given satellite is not visible at all, but is known to be at a certain position in the sky, then you can infer that a building must be blocking the view, and constrain the options for the position fix to only be in the “shadow” of buildings for that line of sight to the satellite.

Supercorrelation™ improves 3DMA methods by cleaning all measurements of multipath interference, and by allowing the receiver to select specific arrival angles to gather measurements from.

The short answer is 'no'. Traditional GNSS signals broadcast in the L1 frequency band are much lower bandwidth signals than the new L5 broadcasts. L5 signals have a correlation peak that is much narrower than L1 signals (29m vs 293m). This means that L5 receivers are limited to coherent multipath interference distortions of around 15m instead of 150m for L1. Supercorrelation™ is still very useful in removing that 15m of error on an L5 signal due to multipath interference.

Further to this, an L5 signal that bounces from a building 50m from you still arrives with 100m of extra path delay regardless of the sharpness of the correlation peak, and will distort your position fix by 100m if not filtered correctly. The best way to filter all the incoming satellite signals is by their angle of arrival using Supercorrelation™, regardless of whether they are L1, L2, or L5 signals.

Spoofing involves broadcasting fake satellite signals in order to confuse the receiver. Spoofing is used by criminal networks, malicious actors and fraudsters. As the cost of spoofing technology comes down, the danger to business and consumers is increasing.

Because Supercorrelation™ can determine the angle of arrival of any signal, it can flag fake signals as spoofers, allowing for them to be removed from position calculations and flagged for further actions, such as reporting to authorities.

The key properties of an antenna are the sensitivity and their directionality. The low cost antennas in smartphones are low sensitivity and omnidirectional - they cannot determine which direction the signal came from. Supercorrelation™ effectively adds a software-based antenna that mimics the capability of a high-powered hardware antenna - boosting the sensitivity by a factor of 10 and is the only way to make a hardware omni-directional antenna capable of determining angle of arrival.

The oscillator is used to measure the frequency that a radio is tuned to - the accuracy with which you can determine this is important for the position accuracy of the GNSS chip.

Oscillators in smartphones are very low cost and low in performance. Supercorrelation™ boosts the performance to the level of a much higher quality oscillator, without the need for additional hardware space, power consumption or cost.

S-GNSS or Synthetic GNSS describes a GNSS receiver that is enabled with a Synthetic Antenna, Synthetic Oscillator and Synthetic TCXO working alongside its hardware components.

Each of these synthetic components are enabled by having Supercorrelation™ integrated into the receiver.

The picture to the right shows a diagram of an S-GNSS receiver.

Supercorrelation™ is designed to mitigate inherent problems with GNSS to improve absolute positioning sensitivity, integrity and accuracy. It operates at the level of the GNSS receiver.

D-Tail is our human motion modelling technology that is designed to construct an accurate relative path through space, using inertial sensor data.

D-Tail can be embedded into our Supercorrelation™ technology for use cases that require human motion modelling, such as for smartphones and smartwatches

Without correction from GNSS-based positioning, D-Tail experiences 1% drift, which means that for every 100m travelled, it gains 1m of error.

D-Tail requires at least accelerometer, gyroscope and GNSS data. It can also use Barometer and Magnetometer data, both of which will improve the quality of the D-Tail output.