For my school project we have to calculate te relative posistion between chair using rssi in a large office building, right now have it gotten to work that the chairs can see each other posistion in rssi value, but we them to see each other in meters.

So after some researche I came to this link: Calculate distance from RSSI

In this link I found the formula: RSSI (dBm )=−10n log10(d)+A

So the problem is I can determine the other variables but I have trouble determining the n variable which is the Path-Loss Exponent. In the link they speak of that the value is somewhere between 2.7 and 4.3 but how can I choose which is best for my case.

  • \$\begingroup\$ What system? - WiFi, BLE, ...? \$\endgroup\$ – Russell McMahon Oct 12 '16 at 16:56
  • \$\begingroup\$ The systeem that I'm using is BLE. \$\endgroup\$ – timvanhooff Oct 13 '16 at 7:20
  • \$\begingroup\$ Just as a note, determining positioning based on RSSI can be quite inaccurate. The path loss exponent changes depending on the surrounding objects and the materials they are made of, or if humans are present (giant water cylinders don't help wireless signals too much). \$\endgroup\$ – Catsunami Jan 31 '18 at 17:08

I assume that you are using BLE or similar.
Be aware that RSSI is a very blunt instrument indeed and needs both understanding and a degree of magic to work well.
There is only so much magic available in an office building and if you expect consistent fine precision you will be disappointed.

If working in air (eg assuming that your office is not underwater or filled with solid rock or polystyrene beads or other dielectric) then n should be constant and determinate. The fact that they have given a typical range is a clue that things are less exact in practice than in theory.

If you are doing this only theoretically and not with practical experiment as well then then the value of n used is not too important as it is almost certain to be wrong.
RSSI (signal strength) readings can be affected by reflections, position of objects in the target space, object motion, line of site or not between nodes, number of nodes, presence of other unrelated signals on the allocated frequency band (or stronger ones out of band), directionality (or not) of antennas, front end overload and intermodulation performance, system signal to noise ratio, ... to give a far from complete list.

RSSI distance measurements almost certainly need to be based on multiple signal attempts processed in some way. Having reference nodes of known distance can help. You can have fixed "beacons" with moving targets, or moving beacons with fixed targets, or some mix. There is a VERY large amount of information online on indoor distance determination using RSSI and some fairly basic searching will turn up more than you are liable to be able to assimilate in a sensible timespan.



Here is an edited subset of some notes I wrote for somebody else a few months ago.
Estimote are very active in this area and looking at what they are doing should be informative:

ESTIMOTE LOCATION development kit (!) A must read. But, it starts like this:

  • Estimote Indoor Location SDK is a set of tools for building precise, blue-dot location services indoors. Just configure Estimote Beacons (no sweat, it’s fully automated), attach them to walls and you’re ready to set up your first location.

    From there, the location’s map is automatically uploaded to Estimote Cloud, and voila! You can now embed it in your own app.

    Location accuracy varies depending on location size, shape, and crowd density. In small rooms, it’s as good as 1 meter, in larger spaces it can be around 4 meters on average. Keep in mind that Indoor Location SDK is still work in progress: we’re constantly improving accuracy, adding new features, and optimizing for bigger venues.

Estimote indoor location video 2m-12 https://www.youtube.com/watch?v=wtBERi7Lf3c

Nearables video 2m (smaller, lower range, lower capability, lower cost.) https://www.youtube.com/watch?v=JrRS8qRYXCQ

Estimote location (scroll down) Existing working install beacons, setup and go system

Estimote You tube videos many many many https://www.youtube.com/results?search_query=estimote

eg Estimote Beacons factory - posted December 2013 56s http://estimote.com/indoor/

Building the next generation of context-aware mobile apps requires more than just iBeacon™ hardware. Developers need smarter software: tools that give them control over proximity and position within a given space, without unnecessary hassle. Estimote Indoor Location does just that. It makes it incredibly easy and quick to map any location. Once done, you can use our SDK to visualize your approximate position within that space in real-time, in your own app. Indoor Location creates a rich canvas upon which to build powerful new mobile experiences, from in-venue analytics and proximity marketing to frictionless payments and personalized shopping.

What is Estimote Indoor Location SDK? Download software free.



iBeacon and other RSSI position location:

While iBeacon is principally intended as an information promulgation system, a number of people have used the BLE transmitters as position/ triangulation based on RSSI and other methods. The Wikipedia page provides a good general overview.


BLE mesh networking


An experienced BLE position location developer and vendor.


BLE - more re beacon based


Another http://kontakt.io/

BLE in GEC lighting fixtures


' Google Scholar on ble position detection.


9783319226880-c2.pdf <- URL needed.

      Indoor Position Detection Using BLE Signals Based on Voronoi Diagram

An Analysis of the Accuracy of Bluetooth Low Energy for Indoor Positioning Applications


Indoor positioning with beacons and mobile devices




Development http://get.openbeacon.org/source/#github


Bluetooth Proximity Tag

Nordic BLS IC / modules



I have many many many variably related references if of interest - but, so does Mr Gargoyle.

eg Every picture tells a story


Most of these many many many stories will be of high relevance.

  • \$\begingroup\$ Thank you for your anwser and all the references!! I will definitely look into them. \$\endgroup\$ – timvanhooff Oct 12 '16 at 12:59

If using a well designed WiFi receiver such as in Laptops, there is a Windows App ( get WiFiInspector-Setup- ) that will display WiFi signal in 1dB resolution with stripchart or numeric display.

Then using Friis Loss to define inverse squared Distance losses in RF, one can make a reasonable conversion from RSSI to dB to distance for line of site with care to avoid wall reflections. from -80 dBm to say -30dBm

enter image description here

Body proximity will also affect near field reflections at low levels. Put router high on plastic chair or table and expect all metal structures and some other materials to reflect signals and cause Ricean Fading nulls.

To expect decent coverage, you will need a high quality router such as DIR-880L Warning: Slightest orientation of a few mm of Laptop can cause peaks and nulls in low level areas due to Ricean fading.

You can then chart range of signals in each room and plot Min/max mean vs distance. using strip chart and walking around zone to be measured.

Typically if RSSI

  • -60 dBm excellent
  • -70 dBm good
  • < -70 dBm marginal and baud rates will be affected from interference
  • <-80 dBm poor and barely enough to communicate
  • WiFi speed and SNR threshold or RSSI level are inversely related so "b" speed 11Mbps needs less signal and "n" speed needs more.
  • \$\begingroup\$ Choice of router and central location of router in hallway near ceiling is probably crucial to good results. Don't expect Friss Laws (pun for Loss) to apply inside a building with so many metal stud reflectors. It will be a mess of reflections with a wide scattering or results. Also don't expect your Rx RSSI to dB calibration to be as good as a laptop with this software. \$\endgroup\$ – Sunnyskyguy EE75 Oct 12 '16 at 13:21

If you want to calculate the path loss exponent 'n' you would first need to train the linear or straight line model. You can do this by determining the path loss for a range of distances within the region of interest. The path loss exponent can then be determined by fitting a straight line to the distance vs path loss data. This can either be done in MATLAB or if you are not good with coding using the linear regression model within excel. More the data used to find 'n' better it is. But remember that the linear model just gives you averaged out results. The results obtained by this approach may vary from reality quite appreciably.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.