Most wireless receiver modules available today have a received signal strength indicator (RSSI) that gives the signal strength measured at the receiver and which is associated with each received signal source. This information is usually part of a packet stream that gives the source identification and RSSI level in power levels of dBm. The RSSI signal is usually a nice diagnostic tool that shows the margin of the received signal compared to the sensitivity of the receiver.

It does also give some indication of how close the source of the signal is, if one has an estimate of the source power. If one does not know the strength of the source, then one or even two RSSI signals is of little use in finding the location or distance to the source, although most triangulation methods such as those that use phased arrays can provide a bearing angle to the source.

However, four networked transceivers arranged in a triangular pattern and equipped with omnidirectional antennas can determine both the transmitter's power level and the Cartesian coordinates (x,y,z) of the transmitter relative to the receivers when they share their RSSI information. A simplified version of this method was first described in patent #7,283,127, issued October 16, 2007, which claims the methods for triangulation of static magnetic fields and briefly mentions the use of a similar algorithm for triangulation with alternating electromagnetic fields.

I also agree that it is unproductive to approach this in a negative fashion - implying that people don't have knowledge of a subject because you disagree with their methods reminds us about who we learned to play nice in kindergarten. Every method has inadequacies - why not just use the good parts of every method to achieve the goals we would want? For instance, if the non-isotropic antenna will cause problems for the algebra at certain angles, shield those problem angles of the antenna to prevent reception of the signal. It is not mentioned in the article that the method would only work with 4pi steradian coverage - hemispherical coverage is an option as well. We live with imperfections in any new method, but that doesn't exclude it from being considered or used as an option or in conjunction with other methods. In other words, there is no one perfect design approach that should replace any others - it all depends on the circumstances as to which approach will work best. Phased arrays have their place but also suffer from interference at the characteristic frequency of the array - these problems can be resolved by combining energy measurement in conjunction with it.

Thanks for the discussion! With my limited knowledge of radio direction finding I only can repeat that to make the author's linear algebra working, requires using perfect omni antennas and allowing no multipath.
In real world, this is a rare case.
Rich is completely correct but his contribution confirms mine in all details.

It is unnecessary to take such combative tones in the discussion. I think we need to encourage discussion of RF methods and techniques in general- old, new, useful and flawed. I believe the limitations of the method were admitted by the author and it was unproductive to insinuate the response was from the author. Quite frankly I would hate to have either of you referee my papers but such is life... We must work within the culture of our profession. Still one could always hope for change.

I understand the author passionately defending his invention.
I would invite him- as he knows that a dipole has 1.76 dB gain - in linear terms which his algebra works- to calculate an angular error of his miraculous system with one dipole having the directivity of those ~50%.
Also, a dipole has a deep minimum in its radiation pattern, a good one has 30 dB, or one thousandth. What author's algebra would do with that?
It may be true I do not know as much about US patenting like the author- I had some 20 Czechoslovak patents, sorry! But I still can see that US patent referees do not know how nature works!
Regards,
Jiri Polivka

This statement is not true - "For its use in "wireless", antennas must be perfectly omnidirectional, otherwise the method is completely flawed." This statement sounds more like the writer is talking about digital logic (or Platonic Logic) rather than RF. Even a dipole antenna only has 1.76 dBi - 1.76 dB above a perfect omnidirecitonal (which of course perfect omni does not exist but is used as 0 dBi for gain calculations) but 1.76 dB is not going to cause that much error in these calculations to render the method "completely flawed", in fact the same signal from many directions will experience this same gain and cancel in the equations. There are the two angles on the dipole (0 and 180 degrees) where the gain goes to zero and one would expect errors in the calculations for the source being in these locations but for the 90% of the of solid-angle coverage the algorithm would still work as expected. Having areas that are excluded is a small price to pay for x,y,z triangulation (and yes I did use the word TRIANGULATION, which from Webster - "any similar trigonometric operation for finding a position or location by means of bearings from two fixed points a known distance apart" - broadly meant to find the LOCATION of the source using triangles - please take another close look at Figure 1 and tell me there aren't angles embedded in the formulas - arcTan [r^2/(x^2 + y^2 + z^2)] ?). Also, everybody knows that any TRIANGULATION method using RSSI can be affected by multi-path, reflections and absorption of materials - this does not stop people from using RSSI for triangulation even in less-than-ideal circumstances.
As far as who invented what and when - Mr. Polivka needs to read up on US patent law which states those who file the patent first are the official inventors (thus, the patent referees are in the clear). Also, Mr. Polivka states that he and Mr. Watson Watt used two sensors in their previous experiments so they invented it first (isn't it amazing how many detractors of an idea like to claim ownership of it?). Two sensors gives you the ratio of y/x in a plane and therefore just the bearing angle - a concept used before WWII, but I would like to see two sensor equations simultaneously solve for the four variables of x,y,z and power level? This is simple linear algebra - number of sensors = number of variables solved, but not so easy that everybody has figured out the geometry of a four-sensor array. In the case of two sensors you are covering for the unknown power level and the angle to the source and that is all. The use of four sensors to solve for four variables to find coordinates in 3-space is the unique part.
As far as the FoxHunts go ... well, that's where I got the idea from! As a radio amateur (KB7WVJ) watching other amateurs use phase angle, rough amplitude estimates and doppler direction finding at FoxHunts, I thought this method would be worth trying. LOL, KB7WVJ

As mentioned in the Introduction, Mr. Harney's "triangulation method" may work with static magnetic fields. For its use in "wireless", antennas must be perfectly omnidirectional, otherwise the method is completely flawed.
Concerning Mr.Harney's patent, in 1920's Mr. Watson Watt (the one who erroneously claimed to be the first inventor of radar -he was not) used a pair of omnidirectional shortwave antennas to locate a lightning strike. At least two antenna pairs were needed to "triangulate" the position.
Antennas generally have directional radiation patterns, and their maxima and preferably minima are used to locate a transmitter. Mr. Harney apparently never read a radio location textbook, or is familiar with radio amateur techniques of "Fox-Hunt"; at HF, magnetic antenna minimum is used, and at UHF, an array maximum, to point to a hidden adio transmitter.
Mr. Harney's method will need perfect omnidirectional antennas and free space to work. Any object reflecting or dispersing a wave fromt will confuse his system.
Some 25 years ago I patented a similar system to locate a point over a 10x10 inch plane. I used an LED as a source, and two pair of light sensors with opamps. To make their patterns omnidrectional, I used plastic light guides. The system worked well, similarly to a PC mouse, only with a smooth response.
So Mr. Harney's patented method is not new, would not work in a real situation. His missing knowledge of radio location principles and techniques is a pity - I am, however, disturbed by the fact that Patent Office referees failed to know how nature works and that similar methods were published many years ago. Also, "triangulation" is about determining angles, not signal levels. In radio location, we use logarithmed levels in decibels to an advantage. Also our senses like hearing works logarithmically.
Jiri Polivka

David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.

To save this item to your list of favorite EE Times content so you can find it later in your Profile page, click the "Save It" button next to the item.

If you found this interesting or useful, please use the links to the services below to share it with other readers. You will need a free account with each service to share an item via that service.