How Far Will Your Ranging Radio Operate?

We have just released our newest version of ranging and networking software (current customers should have received a download link via email).  RCM 2.8 and RangeNet 1.3 both have significant improvements.  I am most pleased with the following:

  • Acquisition algorithm improvements now allow us to acquire reliably at lower SNRs.  This is about a 2-3 dB improvement.  Your quality of service will be better and your operating ranges longer.
  • The leading edge detection (LED) algorithm has been improved and we now have sub-centimeter accuracy when operating in direct line of sight.
  • Operation indoors through walls is also improved.
  • The number of communications channels has been increased from 7 to 11.

For a complete list of changes check out the release notes that accompany the software.

In the process of testing our acquisition improvements, we operated at extreme distances.  This is always interesting because it starts to touch on the popular question of “how far will your radio operate?”

The routine answer to that question is “it depends.”  While this is accurate, it is also unsatisfying.  Since this release increases operating range, I thought I would take this opportunity to discuss this subject in a bit more detail.

First of all, UWB technology operates at high enough frequencies that the RF energy does not bend around corners.  RF will also not operate through large metallic or conductive structures like buildings or metal fences.  RF will travel through many (but not all) types of building materials, but in doing so it will normally be attenuated – sometimes by just a little bit and other times by quite a lot.  So if you want to go really far, you need to find an open area.  Like out in the country or in a large park.

Second, Time Domain performance specification sheets are written by conservative engineers and we tend to build margin into our distance claims.  When we quote a range we mean it.  Your performance will be better than what is in the spec.

Third, performance can also depend on the level of interference you might be getting from other RF services.   You really shouldn’t expect much interference but we transmit only 0.000050 watts while other transmitters operate at 5.000000 watts (or higher).  While these other transmitters are in different frequencies (and so we shouldn’t see interference), those transmitters always splatter a bit of energy our way.  When measuring, it is good to know if you are seeing interference.

Fourth, the terrain over which you operate will affect performance.   There are a couple of mechanisms.  The Fresnel effect will sometimes help and sometimes hurt performance.  Fresnel effects are the result of the addition of energy from the direct path with energy bouncing off of reflectors, like the ground.  Sometimes the energy is in phase and constructively adds, and sometimes it is out of phase and cancels.  The natural roll of the ground is also a big factor.  It isn’t very evident over 10s of meters but over 100s of meters you can have enough variation in elevation to block your signal.  Changes in elevation will also change how Fresnel helps or hurts your transmissions.

Fifth, a bit of antenna gain always helps.

I found two examples that illustrate these effects.

The first example illustrates performance on a quiet country road.  There are several things of note.

  • The blue line is a measure of the noise in the environment.  The magnitude of that signal doesn’t change with distance.  In fact, it is equal to what you would see if you were operating a conducted test over cables.  This is an RF quiet place.
  • We collected the data using PII 8.  The link operated out to 600 meters and then it started to get a bit intermittent.   A quick look at our data sheet will indicate that PII 8 has a maximum range of 177 meters.    Yes, we are conservative but…
  • We are also getting benefit from Fresnel effects.  You can see this effect in two ways.   At 50 and 93 meters there is a dip in signal strength.  This is the result of Fresnel cancellation.  Between 100 and 300 meters, the SNR does not change with distance.  It should drop as 1/r2, but it doesn’t because Fresnel is constructively increasing the signal strength.   Past 300 meters the SNR is more a factor of the changes in elevation of the surface of the ground.  The acquisition improvements are a bit harder to see.  With 2 dB more margin, the Quality of Service from 300 to 600 meters is improved by about an order of magnitude and the link will operate about 20% further.



The second example was taken in a large suburban park. In this test we used a simple back reflector to increase the gain of the system by about 5 dB per antenna.




There is a lot of interesting information in this data.

First, notice that the noise (blue line) is not flat.  In this test, the responder is parked near a giant cell tower (populated with dozens of antennas) and the requester logs data as it drives away from the responder.  When the units are close to each other, the noise is about 7 dB higher than when the requester is 1400 meters away.  At 1400 meters there is virtually no noise.  However, since the link needs both radios to operate, the link operating range is set by the noise measured in the worst possible location.

Second, you can more easily see the effect of changes in elevation.  From 550 to 1200 meters the terrain drops into a shallow valley about 4 meters deep.  At 1200 meters the terrain rises by about 6 meters.  At 1300 meters the terrain starts to drop and by 1440 meters the radios lose line of sight.

The effect of the back reflector is also obvious.  Compare the SNR of the two examples at 400 meters.  In the first case the SNR is 22 dB, in the second case the SNR is 32 dB.  Two 5 dB antennas help a lot.

Finding nice open areas is rare.  As a consequence, these examples might not be relevant to many users.  If your application is limited to shorter operating distances, then you have lots of options because operation at long ranges means successful operation at short ranges but with LOTS of margin.  Margin is good.  You can use this margin to operate more robustly at short ranges or you can reduce the PII and trade range for update rate.

Leave a Reply