[Bluez-devel] multi-frequency scanning & possible contract work

[Bluez-devel] multi-frequency scanning & possible contract work

[Paul Huber]

[-- Attachment #1.1: Type: text/plain, Size: 2039 bytes --]

I need to use a directional antenna focused on an area about 6 meters x 6
meters that can scan for Bluetooth mobile phones as they pass through the
area, and identify them very quickly.  It's for a new startup business
that's focusing on a retail application running on an embedded Linux box.

The problem with the usual usb Bluetooth dongle and a normal scan is that in
a busy area with hundreds of Bluetooth phones nearby, the scan can take a
very long time, like several minutes, and will pick up lots of phones of
people all over the store.  We just want to detect them when they come in
through the entrance.  So, we're going to use a directional antenna that
focuses just on a 6 meter x 6 meter space in front of the entrance so it
doesn't pick up phones outside that area.

However there's still a problem that a traditional Bluetooth scan takes 15+
seconds to pickup devices, and that's too long.  The shopper may pass
through the detection area in only a couple seconds.  So we need to be able
to do the scans very quickly.

I don't know that much about the underlying technology, but I understand
that a typical Bluetooth scan consists of scanning across several
frequencies and hopping from frequency to frequency to find phones on each
frequency.  Therefore, my question is this: instead of having 1 bluetooth
dongle that's hopping across several frequencies, can we hook up multiple
bluetooth dongles to the same antenna, and have each one locked onto just
one frequency to make the process go faster?  In other words, can we balance
the 15 second scanning process across 5 Bluetooth dongles so we can do the
same scan in only 3 seconds?

If not, does anybody else have any suggestions for how to accomplish the end
goal of being able to very rapidly detect phones as they pass through an
area?

We do have a little bit of funding, so if something like the multi-dongle
approach method i mentioned works, and it requires some special coding or
mods to bluez, we could compensate a developer for helping out with this.

Thanks

[-- Attachment #1.2: Type: text/html, Size: 2155 bytes --]

[-- Attachment #2: Type: text/plain, Size: 314 bytes --]

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

[-- Attachment #3: Type: text/plain, Size: 164 bytes --]

_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] multi-frequency scanning & possible contract work

[Brad Midgley]

Paul

Are you using hcitool from the command line? Use the D-Bus bindings
(eg python) with a listener method for discovered devices and you'll
get piecemeal results that are likely more timely. Maybe look into
extended inquiry too.

On 10/30/07, Paul Huber <paul2004x@gmail.com> wrote:
> I need to use a directional antenna focused on an area about 6 meters x 6
> meters that can scan for Bluetooth mobile phones as they pass through the
> area, and identify them very quickly.

-- 
Brad

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] multi-frequency scanning & possible contract work

[Paul Huber]

[-- Attachment #1.1: Type: text/plain, Size: 574 bytes --]

Brad,

I'm using the C libraries directly within my own code, so I do get the
piecemeal results, and have played with extended inquiry.  But I find that
it's still not fast enough, which was why I was hoping to try simultaneous
scanning with multiple dongles, or something like that to speed it up.

Do you know if bluez has enough low-level control over the usb device to do
something like that?  ie have two devices scanning different frequencies?
Or does the usb interface just have a high level 'start scanning' without
the low-level control that would be needed?

Paul

[-- Attachment #1.2: Type: text/html, Size: 634 bytes --]

[-- Attachment #2: Type: text/plain, Size: 314 bytes --]

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

[-- Attachment #3: Type: text/plain, Size: 164 bytes --]

_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] multi-frequency scanning & possible contract work

[Matthias Becker]

[-- Attachment #1.1: Type: text/plain, Size: 2513 bytes --]

A scan takes about 10 seconds. Only the name resolution takes that long.
Names can be resolved in parallel using many adapters.

 

Von: bluez-devel-bounces@lists.sourceforge.net
[mailto:bluez-devel-bounces@lists.sourceforge.net] Im Auftrag von Paul
Huber
Gesendet: Dienstag, 30. Oktober 2007 21:06
An: bluez-devel@lists.sourceforge.net
Betreff: [Bluez-devel] multi-frequency scanning & possible contract work

 

I need to use a directional antenna focused on an area about 6 meters x
6 meters that can scan for Bluetooth mobile phones as they pass through
the area, and identify them very quickly.  It's for a new startup
business that's focusing on a retail application running on an embedded
Linux box. 

The problem with the usual usb Bluetooth dongle and a normal scan is
that in a busy area with hundreds of Bluetooth phones nearby, the scan
can take a very long time, like several minutes, and will pick up lots
of phones of people all over the store.  We just want to detect them
when they come in through the entrance.  So, we're going to use a
directional antenna that focuses just on a 6 meter x 6 meter space in
front of the entrance so it doesn't pick up phones outside that area. 

However there's still a problem that a traditional Bluetooth scan takes
15+ seconds to pickup devices, and that's too long.  The shopper may
pass through the detection area in only a couple seconds.  So we need to
be able to do the scans very quickly. 

I don't know that much about the underlying technology, but I understand
that a typical Bluetooth scan consists of scanning across several
frequencies and hopping from frequency to frequency to find phones on
each frequency.  Therefore, my question is this: instead of having 1
bluetooth dongle that's hopping across several frequencies, can we hook
up multiple bluetooth dongles to the same antenna, and have each one
locked onto just one frequency to make the process go faster?  In other
words, can we balance the 15 second scanning process across 5 Bluetooth
dongles so we can do the same scan in only 3 seconds? 

If not, does anybody else have any suggestions for how to accomplish the
end goal of being able to very rapidly detect phones as they pass
through an area?

We do have a little bit of funding, so if something like the
multi-dongle approach method i mentioned works, and it requires some
special coding or mods to bluez, we could compensate a developer for
helping out with this. 

Thanks


[-- Attachment #1.2: Type: text/html, Size: 4932 bytes --]

[-- Attachment #2: Type: text/plain, Size: 314 bytes --]

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/

[-- Attachment #3: Type: text/plain, Size: 164 bytes --]

_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] multi-frequency scanning & possible contract work

[Peter Wippich]

Dear, 

what Paul likes to do is a fast inquiry. The time a Bluetooth device does 
an Inquiry can be set by an HCI command. So first of all I think he should 
use the HCI layer directly, not the libraries on top of it.

If you want to use multiple dongles there is a concurrency problem. 
In theory it should be possible to synchronize multiple dongles, but in 
practice you need access to the baseband / link manger layer, which you 
usualy don't have. It is part of the dongle firmware and this is not 
disclosed by the chip manufacturers. 

You can also build your own RF / basband Hardware and write a specialized 
baseband firmware for this. Don't name the resulting device Bluetooth or 
use the Bluetooth Logo on it because it will for sure not be possible to 
get such a beast through Bluetooth qualification. 

However, this stuff is completly off topic and not related to BlueZ at 
all. 

Ciao, 

Peter 

On Wed, 31 Oct 2007, Matthias Becker wrote:

> A scan takes about 10 seconds. Only the name resolution takes that long.
> Names can be resolved in parallel using many adapters.
> 
>  
> 
> Von: bluez-devel-bounces@lists.sourceforge.net
> [mailto:bluez-devel-bounces@lists.sourceforge.net] Im Auftrag von Paul
> Huber
> Gesendet: Dienstag, 30. Oktober 2007 21:06
> An: bluez-devel@lists.sourceforge.net
> Betreff: [Bluez-devel] multi-frequency scanning & possible contract work
> 
>  
> 
> I need to use a directional antenna focused on an area about 6 meters x
> 6 meters that can scan for Bluetooth mobile phones as they pass through
> the area, and identify them very quickly.  It's for a new startup
> business that's focusing on a retail application running on an embedded
> Linux box. 
> 
> The problem with the usual usb Bluetooth dongle and a normal scan is
> that in a busy area with hundreds of Bluetooth phones nearby, the scan
> can take a very long time, like several minutes, and will pick up lots
> of phones of people all over the store.  We just want to detect them
> when they come in through the entrance.  So, we're going to use a
> directional antenna that focuses just on a 6 meter x 6 meter space in
> front of the entrance so it doesn't pick up phones outside that area. 
> 
> However there's still a problem that a traditional Bluetooth scan takes
> 15+ seconds to pickup devices, and that's too long.  The shopper may
> pass through the detection area in only a couple seconds.  So we need to
> be able to do the scans very quickly. 
> 
> I don't know that much about the underlying technology, but I understand
> that a typical Bluetooth scan consists of scanning across several
> frequencies and hopping from frequency to frequency to find phones on
> each frequency.  Therefore, my question is this: instead of having 1
> bluetooth dongle that's hopping across several frequencies, can we hook
> up multiple bluetooth dongles to the same antenna, and have each one
> locked onto just one frequency to make the process go faster?  In other
> words, can we balance the 15 second scanning process across 5 Bluetooth
> dongles so we can do the same scan in only 3 seconds? 
> 
> If not, does anybody else have any suggestions for how to accomplish the
> end goal of being able to very rapidly detect phones as they pass
> through an area?
> 
> We do have a little bit of funding, so if something like the
> multi-dongle approach method i mentioned works, and it requires some
> special coding or mods to bluez, we could compensate a developer for
> helping out with this. 
> 
> Thanks
> 
> 


|       Peter Wippich                   Voice: +49 30 46776411          |
|       G&W Instruments GmbH            fax:   +49 30 46776419          |
|       Gustav-Meyer-Allee 25, Geb. 12  Email: pewi@gw-instruments.de   |
|       D-13355 Berlin  / Germany                                       |


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] multi-frequency scanning & possible contract work

[Jim Carter]

On Tue, 30 Oct 2007, Paul Huber wrote:

> I need to use a directional antenna focused on an area about 6 meters x 6
> meters that can scan for Bluetooth mobile phones as they pass through the
> area, and identify them very quickly.  It's for a new startup business
> that's focusing on a retail application running on an embedded Linux box.

Since you're using non-provincial units I assume you're in the European 
Union, but in the USA there's a specific reg forbidding cellphone spam.  
Pro-actively dialling the customer's phone could well give a hostile 
impression.  A customer-initiated interaction such as WAP is a lot more 
polite, but the shopkeeper would want to make it available to everyone 
including those looking through the window.

If the technical and regulatory issues can be straightened out, it would be 
very valuable to both the customer and the business to provide a product 
search feature, tech specs for products, etc.  All the advantages of online 
shopping, but you can look at the physical product and buy it on the spot.

James F. Carter          Voice 310 825 2897    FAX 310 206 6673
UCLA-Mathnet;  6115 MSA; 405 Hilgard Ave.; Los Angeles, CA, USA  90095-1555
Email: jimc@math.ucla.edu    http://www.math.ucla.edu/~jimc (q.v. for PGP key)

-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] multi-frequency scanning & possible contract work

[Marcel Holtmann]

Hi Paul,

> I need to use a directional antenna focused on an area about 6 meters
> x 6 meters that can scan for Bluetooth mobile phones as they pass
> through the area, and identify them very quickly.  It's for a new
> startup business that's focusing on a retail application running on an
> embedded Linux box. 

the use of an directional antenna with a Bluetooth chip is illegal as
far as I know. You really wanna check the regulations here.

> The problem with the usual usb Bluetooth dongle and a normal scan is
> that in a busy area with hundreds of Bluetooth phones nearby, the scan
> can take a very long time, like several minutes, and will pick up lots
> of phones of people all over the store.  We just want to detect them
> when they come in through the entrance.  So, we're going to use a
> directional antenna that focuses just on a 6 meter x 6 meter space in
> front of the entrance so it doesn't pick up phones outside that area. 

You can use the RSSI value to pick only the phones in range. This needs
some manual tuning, but works pretty well so far. And if you simply not
use the name resolving, then you will see immediately devices in range
and also devices passing by. The bluetooth-proximity example from the
bluez-gnome package does this.

> However there's still a problem that a traditional Bluetooth scan
> takes 15+ seconds to pickup devices, and that's too long.  The shopper
> may pass through the detection area in only a couple seconds.  So we
> need to be able to do the scans very quickly. 
> 
> I don't know that much about the underlying technology, but I
> understand that a typical Bluetooth scan consists of scanning across
> several frequencies and hopping from frequency to frequency to find
> phones on each frequency.  Therefore, my question is this: instead of
> having 1 bluetooth dongle that's hopping across several frequencies,
> can we hook up multiple bluetooth dongles to the same antenna, and
> have each one locked onto just one frequency to make the process go
> faster?  In other words, can we balance the 15 second scanning process
> across 5 Bluetooth dongles so we can do the same scan in only 3
> seconds? 

If you keep calling it scanning, then you have no clue how Bluetooth
actually works. So you might wanna get one of the Bluetooth books that
describe the basics behind this technology. And your math will not work
out that easily. It is not that simple.

> We do have a little bit of funding, so if something like the
> multi-dongle approach method i mentioned works, and it requires some
> special coding or mods to bluez, we could compensate a developer for
> helping out with this. 

Just a warning up-front. If you don't use the D-Bus API and wanna use
the HCI raw socket, you have to make your code public and put it under
GPL license.

Regards

Marcel



-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] multi-frequency scanning & possible contract work

[Peter Wippich]

Hi Marcel, 

> the use of an directional antenna with a Bluetooth chip is illegal as
> far as I know. You really wanna check the regulations here.

I don't think so. What you have to observe is that the maximum isotropic 
radiation does not exceed +20dBm. So if you use an directional antenna 
with a gain of, lets say +16 dBm (a somehow typical value) than the 
maximum transmit power that can be used is +4 dBm. And you have to 
convince the regulartory body that this will not be exceeded  under all 
circumstances.....

Ciao, 

Peter 


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel