SparkFun Forums 

Where electronics enthusiasts find answers.

All things pertaining to wireless and RF links
By pekmezcim
#42414
Hi,

I am trying to implement an indoor location detection system using RF modules. I already have Nordic nRF2401 modules. So i come up with various of algorithms based on measuring the distance between two RF modules.

So i simply try different ways to measure it but the result was not much more than being in range or not for different output powers. (there are only 4 different output power option and even the smallest output power causes too much range for me)

I found out an option for me to replace the ceramic antenna with some wire and resistors but I am not sure about the length of the wire and the resistor value.

Another option seems like using RSSI pins and measure the distance between the RF modules where my module(RF2401) does not provide any RSSI output. Maybe changing the RF module would help but not sure about which one to use.

So I am kinda stuck here. It would be great to know wheather any of these options would work or not for distance measurement.
And which module to use for RSSI measurement or how to design an antenna for a specific range.

Cheers,

Mehmet
User avatar
By leon_heller
#42424
There are techniques for determining distance from signal strength but they are very complex, involving networks of microcontrollers and RF transceivers. You can't simply use an RSSI output.

Leon
By stevech
#42428
distance measurements using RSSI are very approximate and realistically, the technique requires 4 or more immobile (fixed) devices to simultaneously receive a moving node's signal. An algorithm, using a database of RSSIs per fixed node for a reference mobile's movements at time of survey, is the essence of the technique. There are 3 or so products on the market that do this, in 900MHz and 2.4GHz/802.11. All such "RF footprinting" techniques are vulnerable to changes in the physical environment altering and invalidating the survey database: big changes in furniture placement, etc. Also, RF attenuation due to the human body and other things will skew the accuracy.

Alternatively, one can use time-difference-of-arrival (TDOA) methods for location estimates. It too needs 3-4+ fixed nodes, each with a costly special receiver to support TDOA and a common time reference. There are a few such products on the market.

The best approach is to use a high density of fixed nodes. This is beginning to be done with 802.15.4/ZigBee meshes. There are retrofit kits to put a cheap ($50 or less) node in a ceiling light fixture or exit sign. So the mobile node's location is often just a radius from the fixed node with the strongest signal, based on some averaging/statistics.

Related to 802.15.4 and perhaps ISO18000, there is a not so useful standard for "Real Time Location Systems" RTLS, in RFID and telemetry systems, as I've described above. For 2.4GHz, e.g., ZigBee, it is ISO/IEC 24730, and it is based on RF footprinting as I recall.
By pekmezcim
#42588
Thanks for your fast and serious replies,

I had a much deeper research and found out that, it is really going to be hard to resolve the locating algorithm within a microcontroller.
Especially normalizing the RSSI signals and reference syncronisation for TDOA.

While searching on google, i spotted to CC2431 module of chipcon. It seems pretty useful and easy.

But still on progress of research. I would be pleased to hear if you have any experience about this module or any other module that enables RTLS.

Cheers,

Mehmet