There are several methods for implementing direction finder.

1. One of the method is by using directional antenna pointing to target device, and reading the received signal strength from the target device. This method also need the device to keep transmitting, The following "direction finder" use such simple directional antenna and is described here: http://www.pki-electronic.com/products/interception-and-monitoring-systems/gsm-direction-finder/

Yet, if this is the method that is used with the pki device, I wander how it can discern the device among other devices, because as I understand, the uplink frequency is shared among several devices nearby.

2. There are other method too, using Pseudo-doppler DF method, such as done by Balint Seeber: http://wiki.spench.net/wiki/SDRDF But I see that the demonstration is done on VHF (not GSM/LTE)

I would like to ask:

  • How does the first method (directional antenna) can find a device among other with same frequency ?
  • Can the 2nd method (SDRDF) be used to locate GSM-frequency mobile device ? and discern a device among several with same frequency ?


  • 1
    $\begingroup$ Have you considered using loop antennas? The nulls on each side are much sharper than any peaks you will find with directional antennas. But they need to rotate like any directional antenna, and they are bidirectional unless you use them with a shied on one side. $\endgroup$
    – SDsolar
    Commented Apr 29, 2017 at 8:09
  • 1
    $\begingroup$ Does it help solving the base problem of multiple devices using the same uplink frequency ? Thx $\endgroup$
    – ransh
    Commented Apr 29, 2017 at 19:47
  • $\begingroup$ Lots of variables involved before that can be answered. Like is the target signal ID-able? If so, then you can find the direction where it disappears using a loop. Also, how far apart are all the other signals? If they are all in a tight cluster then the problem is moot, I suppose. In any case, the null of a loop is much sharper than either directional antennas or TOA (time-of-arrival) like the police use on their cars (with 4 antennas) to find money packs - They only get an 8-direction display. $\endgroup$
    – SDsolar
    Commented Apr 30, 2017 at 0:32
  • 1
    $\begingroup$ @ransh, sorry, your GSM-related question is about a specific radio technology that is not amateur radio, and hence is off-topic here; SDSolar and the others will help you answer it in a general manner. But for GSM-specific direction finding: I think you're answering your own question. With purely directive-antenna power detection, no, you can't tell different handsets apart; you already know (and have been told) that. $\endgroup$ Commented Apr 30, 2017 at 13:52
  • 1
    $\begingroup$ So, the other question "can it be done": yes, "how can it be done for GSM": Off-topic here. $\endgroup$ Commented Apr 30, 2017 at 13:54


Browse other questions tagged .