Re: Indoor Geolocation USING WIFI AND TRILATERATION TECHNIQUE

[Paul Gardner-Stephen]

Hello,

This is a noble undertaking, but one that is very hard because all of the options have difficulties.

Some of the options and challenges that I am aware of include:

1. Translateration (using time of arrival to estimate distance) will be practically impossible with Wi-Fi, as you will almost certainly not be able to get the arrival time of packets at nano-second resolution.  Remember that light travels 300m in one micro-second, so even with 1ns accuracy you will be limited to 0.3m accuracy.

2. RSSI (received signal strength) can be used, but will be confounded when there is no line of sight between devices. If you are thinking of using phones, you also have to bear in mind that the Wi-Fi antennae in most newer phones do not have uniform isotropic power.  That is, they are partly directional.  You would need to know the orientation of the phone.  Then you have proximity to the human body and all sorts of other nasty problems to deal with.

3. You might be able to use the FM radio receiver to get FM signal phase (and ideally difference in phase between different FM radio stations), but I suspect that the FM radio in most phones can't supply this information in any useful way.

4. Multiple-device RSSI readings is perhaps the most feasible, where you measure the signal strength of a Wi-Fi packet from the perspective of several receives.  You could then come up with a model that starts to resolve the relative location constraints on each device.  As you begin to constrain the location of each device, the locations of the remaining devices become easier to resolve.  Combine that with dead-reckoning using the in-phone accelerometers so that you can get tracks of movement combined with RSSI, and might just be able to build a reasonably accurate model. You would probably need root access and a Wi-Fi chipset that supports monitor mode to develop and test the solution.

My suggestion if you still really want to pursue this, is to explore (4).  If you can crack it, you will make a significant contribution to the state of the art.  I would be willing to provide some input, advice and testing if you are willing to make the results open-source.

It would also potentially be the kind of project to undertake as a PhD (if you aren't already doing that ;)

Paul.

On Thursday, May 2, 2013 10:34:43 PM UTC+9:30, hope...@gmail.com wrote:

Hi everyone .I am actually working about Indoor Geolocation via wifi . and I am using Android java for programming . any idea or some tutorials or source code, I need some examples to better understand .Is there any applications about this subject ??  thank you :)

[Kristopher Micinski]

I agree with Paul.

I've done something like this before using Android devices and RSSI
information: RSSI correlates so horribly with distance that the
statistical significance of predictions was so bad it was completely
unusable. This was based off of RSSI readings from four different
devices.

I've known a few people who have done this: all of them just map the
locations of wifi APs, look for the one that's closest, and say
"you're around this point."

I don't know how much material would be left for a PhD, my
understanding was that to do this accurately you combine a bunch of
prediction algorithms. This is already commodity enough that there
are a few startups doing this:

http://www.navizon.com/product-navizon-indoor-triangulation-system

Kris

> --
> You received this message because you are subscribed to the Google Groups
> "android-platform" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to android-platfo...@googlegroups.com.
> To post to this group, send email to android-...@googlegroups.com.
> Visit this group at http://groups.google.com/group/android-platform?hl=en.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Terrance Sullivan]

I see that the original query was to Trilateration or more broadly Multilateration both of which are 3 dimensional computations where triangulation is 2D (not sure what "translateration" is Paul :).

Checkout http://locata.com/ for well established technology that already does multilateration in WiFi (which I took you to mean 2.4GHz ISM band) which by the way was developed at UNSW Paul and at high-resolution so I wouldn't quite claim "practically impossible" quite yet..

/T

[Terrance Sullivan]

Do what, exactly ? 

/T

On 8 May 2013 05:15, <hope...@gmail.com> wrote:

Hi , thanks everyone  for the responses and suggestions, I'm up to do it :) I will explore the (4) .

[Marcelo Caiuby]

Hello

I want to do graduation project about indoor geolocation using RSSI, and I intend build this with the module microchip  MRF24WB0MA which I have, but I'm having trouble with development of hardware, in Brazil is difficult to find this modules to buy and i have just the IC with minimal plate and construct the daugther board is difficult to me.  I intend to use trianulanting algorithm and statistical approach. I would like to know if you could give me some advices about this kind of project and principally , what type of hardware would be more easily to me construct this project
sorry abaout my bad english

[Kristopher Micinski]

This doesn't seem to have anything to do with Android, is there an Android specific question in this?

Kris

--

You received this message because you are subscribed to the Google Groups "android-platform" group.
To unsubscribe from this group and stop receiving emails from it, send an email to android-platfo...@googlegroups.com.
To post to this group, send email to android-...@googlegroups.com.

Visit this group at http://groups.google.com/group/android-platform.

[Howard Zhou]

Hi, I'm also doing sth. about Indoor Localization. I'm wondering about using Channel Frequency Response. It seems that only Rssi is available in Android API. Any idea about how the Rssi is calculated? It is possible to get CFR on Android platform?

在 2013年5月2日星期四UTC+8下午9时04分43秒，Amel Hope写道：

[Kristopher Micinski]

iirc RSSI isn't calculated on Android, it's just read off the chipset.

Kris

> --
> You received this message because you are subscribed to the Google Groups
> "android-platform" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to android-platfo...@googlegroups.com.
> To post to this group, send email to android-...@googlegroups.com.
> Visit this group at http://groups.google.com/group/android-platform.

> For more options, visit https://groups.google.com/d/optout.