wireless hack for smart phone

[ioAlan]

Hi lhs hackers. I run a software dev Co. called ioWorld and have a
project requiring a hardware hack that's outside the scope of my
capabilities; I want to create a wireless fob that a smart phone
(Android, to start with) can detect at approx. 100 cm. Any protocol
(inc. Bluetooth) will be ok for the POC/prototype but there is a
strong commercial application for this tech, so ultimately a lo-power
protocol will be preferable.

If you've hacked something similar previously, or know anyone else
that has, I'd be grateful to hear from you and give you all the
details. Please contact me at either: al...@ioWorld.com or on
07930667390

[Toby Catlin]

Any kind of phone?

NFC sounds like a solution

http://en.wikipedia.org/wiki/Near_field_communication 

Problem is there aren't many phones that support it

http://en.wikipedia.org/wiki/List_of_NFC_enabled_handsets 

t

[Mark Steward]

NFC won't travel 100cm without magic.  Unfortunately, Bluetooth is probably the best bet, but security's something to consider.

Mark

[Toby Catlin]

sorry i misread that to be 10cm

[alan hodes]

Hi - firstly apologies for taking so long to reply (my gmail account is not my primary) - In answer to your question: yes, any type of smartphone but for the sake of the prototype, I expect Android will probably be easiest. As for protocols: will NFC work at 100cm if the fob is active?

Alan

[alan hodes]

I believe that your smartphone can be adapted to have NFC with a modified SIM:

http://www.sdid.com/products1010.shtml

http://www.sdid.com/products1020.shtml

http://bit.ly/HSUvbW

On Fri, May 4, 2012 at 2:00 PM, Toby Catlin <to...@korfball.com> wrote:

[alan hodes]

will NFC work at 100cm if the fob is active?

[Jasper Wallace]

On Wed, 9 May 2012, alan hodes wrote:

> will NFC work at 100cm if the fob is active?

Almost certainly not.

Can the phone be activly scanning for the tag in some way? If so you could
use qrcodes and have the phones user point the phone at them.

> On Fri, May 4, 2012 at 2:01 PM, Mark Steward <marks...@gmail.com> wrote:
> NFC won't travel 100cm without magic.  Unfortunately, Bluetooth is probably the best bet, but security's
> something to consider.
>
> Mark
>
>
> On Fri, May 4, 2012 at 2:00 PM, Toby Catlin <to...@korfball.com> wrote:

> Any kind of phone?NFC sounds like a solution

> http://en.wikipedia.org/wiki/Near_field_communication 
>
> Problem is there aren't many phones that support it
> http://en.wikipedia.org/wiki/List_of_NFC_enabled_handsets 
>
> t
>
> On 4 May 2012 13:09, ioAlan <alanh...@gmail.com> wrote:
> Hi lhs hackers. I run a software dev Co. called ioWorld and have a
> project requiring a hardware hack that's outside the scope of my
> capabilities; I want to create a wireless fob that a smart phone
> (Android, to start with) can detect at approx. 100 cm. Any protocol
> (inc. Bluetooth) will be ok for the POC/prototype but there is a
> strong commercial application for this tech, so ultimately a lo-power
> protocol will be preferable.
>
> If you've hacked something similar previously, or know anyone else
> that has, I'd be grateful to hear from you and give you all the
> details. Please contact me at either: al...@ioWorld.com or on
> 07930667390
>
>
>
>
>
>

--
[http://pointless.net/] [0x2ECA0975]

[Nigel Worsley]

> I believe that your smartphone can be adapted to have NFC with a modified SIM:
> http://www.sdid.com/products1010.shtml
> http://www.sdid.com/products1020.shtml
> http://bit.ly/HSUvbW

Nope. The first two are based on full size SD cards, I rather doubt that any phone
using those is still made.

The third link refers to needing to get a new SIM to make use of the NFC hardware
that the phone already has.

If a phone doesn't have NFC built in then there is no convenient way to add it. For
the latest Android devices it should be possible to use USB to connect to an NFC
interface, but that is rather messy.

> will NFC work at 100cm if the fob is active?

Possibly, it depends on the allowable size of the tag as the antenna size has a big
effect on range. An RFID reader that I was working on a few years ago had about
5cm range with a keyfob and 10cm with a credit card sized tag. Using a 25cm
coil in the reader extended that to 30cm, active tags (which I didn't have available)
should do considerably better.

Nigel Worsley

[Dave Ingram]

On 09/05/12 14:25, Jasper Wallace wrote:
> On Wed, 9 May 2012, alan hodes wrote:
>
>> will NFC work at 100cm if the fob is active?
> Almost certainly not.
>
> Can the phone be activly scanning for the tag in some way? If so you could
> use qrcodes and have the phones user point the phone at them.

That sounds very battery-intensive, so wouldn't seem to suit the OP's
low-power requirement -- also not hugely effective at 100cm distance
without large codes :-)


D

[alan hodes]

I expect then battery life in the fob could be an issue if boosted to circa 100cm range? In any case, if you're interested in this, would you like to see the full spec and the intended application for the use of this tech?

[Mark Steward]

Use Bluetooth.  It comes built into phones, and is designed for low power.  If you really want to make something, use Zigbee.

Mark

[alan hodes]

Does  Zigbee need to be installed on a smartphone or is it something that they already have? I've looked up what  Zigbee is, and from the description (low power, short range, secure etc) it would seem ideal for the intended application.

[Mark Steward]

Zigbee's a lower power and simpler protocol to Bluetooth - it's more suited to simple sensors than, say phones or computers.  Phones don't speak it, though, so you'll need to build a device that can interface directly with the phone (SD card/audio socket/Bluetooth) as well as a corresponding Zigbee board on the fob.

Mark

[tom]

I'd go with bluetooth, we found some cheap 3.3v modules on ebay a while back, i think they were about 6 quid each in total

[Toby Catlin]

oooh I wanted to make some small magnetic bluetooth temperature sensors but was put off by the price of arduino bluetooth shields. Do you have a link for anymore?

thanks

toby

[Mark Steward]

These? http://www.dealextreme.com/p/wireless-bluetooth-rs232-ttl-transceiver-module-80711


Mark

[Mark Steward]

Either make your own breakout:

  http://www.flickr.com/photos/fucksocks/5752228871/in/photostream

Or pay a little extra to get a header board:

  http://www.ebay.co.uk/itm/300696517759

Mark

[Nicholas FitzRoy-Dale]

I've used those. They work well, and can be used without a breakout (you just need rx, tx, vss and vdd -- the hardest part was making a voltage divider for tx on an otherwise 5v circuit).

Nicholas

[alan hodes]

As good as the protocol seems, it sounds too complex - at least for the sake of building a prototype. 

[Nigel Worsley]

> I expect then battery life in the fob could be an issue if boosted to circa 100cm range?

Possibly, but not necessarily. All the tag needs to do most of the time is listen for a carrier,
the main receiver can then be powered up to see if there is any real data being sent. If so,
send the appropriate reply and wait for the reader to disappear before going back to sleep.

> In any case, if you're interested in this, would you like to see the full spec and the
> intended application for the use of this tech?

Yes please, that would make it much easier to suggest the right solution.

Nigle

[Nigel Worsley]

>> Zigbee's a lower power and simpler protocol to Bluetooth

> As good as the protocol seems, it sounds too complex - at least for the sake of building a prototype.

It is, simpler than Bluetooth isn't the same thing as simple! The underlying radio format is IEEE 802.15.4
which does almost all of what you will need for just getting the ID of a tag. Range is a bit naff, but
that is probably an advantage for your application.

Nigle

[alan hodes]

I'll contact you off-group

[Aaron Oliver-Taylor]

Hello hackers,

I have two suggestions regarding this that will permit close range proximity measurements, the first is to use an audible tone, and the second is a simple, RF transmitter/receiver pair operating on one of the lower ISM bands such as the one around 433MHz.  In the first case a [reasonably] high pitched sound could be emitted by the fob (high enough as to be quite unnoticable), just loud enough to be detected with a smartphone's microphone within 1m.  This would only really work in the open air though.  The second option would require some specialist hardware - the simplest, most compatible way would be to downmix the RF carrier to an audio frequency voltage, which is then sampled by the microphone jack.  Once again the power of the RF transmitter could be titrated so that it only works over a few metres, and because there is an inverse square law for EM signals the distance could probably be quite accurately judged.  The advantages of using 433MHz are that the power requirements are much lower than 2.4GHz (lower frequency EM waves carry/require less energy), and they generally penetrate walls etc far better.  You could even have the transmitting dongle send a series of bytes to identify itself every so often (for example every 10 seconds), and track multiple devices with the same receiver.

Aaron

__________________________________________
Aaron Oliver-Taylor
Email: aaron.oli...@gmail.com
Web: aayotee.wordpress.com

[Adrian Godwin]

On Thu, May 10, 2012 at 8:07 AM, Aaron Oliver-Taylor
<aaron.oli...@gmail.com> wrote:
> Hello hackers,
>
> I have two suggestions regarding this that will permit close range proximity
> measurements, the first is to use an audible tone, and the second is a
> simple, RF transmitter/receiver pair operating on one of the lower ISM bands
> such as the one around 433MHz.  In the first case a [reasonably] high
> pitched sound could be emitted by the fob (high enough as to be quite
> unnoticable), just loud enough to be detected with a smartphone's microphone
> within 1m.  This would only really work in the open air though.  The second
> option would require some specialist hardware - the simplest, most
> compatible way would be to downmix the RF carrier to an audio frequency
> voltage, which is then sampled by the microphone jack.  Once again the power
> of the RF transmitter could be titrated so that it only works over a few
> metres, and because there is an inverse square law for EM signals the
> distance could probably be quite accurately judged.  The advantages of using
> 433MHz are that the power requirements are much lower than 2.4GHz (lower
> frequency EM waves carry/require less energy), and they generally penetrate
> walls etc far better.  You could even have the transmitting dongle send a
> series of bytes to identify itself every so often (for example every 10
> seconds), and track multiple devices with the same receiver.
>

The problem you'll have with this (as I've found with some commercial
medium-range RFID tags) is that attenuation of RF is predictable in
free space but affected hugely by solid bodies. So it's difficult to
distinguish between a tag that's out of range and one that's in range
but on the wrong side of the wearer's body.

Depending on your application, you might be able to get around this by
noting the occasional glimpses of a nearby (but shielded) tag, but
this makes low power difficult as you need to scan often.

-adrian

[Aaron Oliver-Taylor]

Good point,  so it will depend on the application as to whether this is a feasible option. The degree of attenuation is a function of frequency,  unfortunately both 433MHz and 2.4GHz will be significantly attenuated by biological tissues (the latter more so),  but lower frequencies,  perhaps 35MHz will be less so.  Also the field type with have an effect,  ie rfid is magnetic.

[alan hodes]

Hi Aaron - these are amazingly well-thought out and creative solutions - especially that you don't know the intended application for the technology. I'll start with the first option: the audible solution - in short, it's impracticable because it will only work in open-air. It also needs to work in restaurants, night clubs, offices - you know: places where people use their phones. Regarding the second option requiring specialist hardware: I start getting concerned about escalating costs to develop it. In fact: the prototype can use ANY protocol just to build a proof-of-concept and have a demonstrable product that will justify mass production. I think that if commercially justifiable (on a mass-produced scale) your suggestion should be investigated taking into account cost/benefit. In any case, I would love to come up to hackspace and show you the full spec for the intended application as I think you might appreciate it. 

On Thu, May 10, 2012 at 8:07 AM, Aaron Oliver-Taylor <aaron.oli...@gmail.com> wrote:

[Mark Steward]

Yeah, better to use a module that does 433MHz for you (for prototyping, the cheap RFM12Bs are good, and we've played with them at the space).  Even if you don't want to share the spec here, some idea of what you're transmitting would be good - is it a signed value like Yubikeys, or would a security layer be needed as well?

[alan hodes]

I'm not sure I understand the question but the fob and the phone would need to be identifying each other as being correctly paired and this should be as secure as possible so as to prevent spoofing. As for sharing the spec here: the patent process has only recently commenced and posting to this forum would constitute "publishing" which would invalidate the application - which is why I'm happy to disclose privately. BTW, Yubikey looks very cool! 

[Ken Boak]

Hi

You need to define your data rate and your intended typical payload.

If you only want to send a few bytes occasionally, then Zigbee or Bluetooth is way overkill in terms of data rate and power consumption.

Personally I'd go with either 433MHz or 868MHz in EU - much better building penetration at lower power than 2.4GHz.

You could use an RFM12B - with which I've had 200m range in a typical suburban street environment.

If you only need <10m range you can get away with much simpler and cheaper transceiver hardware.  Wireless doorbell hardware is perfectly adequate, low cost, low power and relatively easy to hack.

I made a 433MHz "wake up"  receiver that operated at  about 10m range and consumed 50uA at 1.5V.  That sort of power can be produced from energy harvesting - or even a small pv cell powered from natural or indoor lighting.

Try Googling "wake-up receivers for wireless sensor networks"

There's also a good article by Eddie Insam on designing super regenerative receivers

http://www.eix.co.uk/Articles/Radio/Welcome.htm

Once you realise that a  superregenerative receiver is virtually an electronically constrained RF amplifier/transmitter, it can lead to some very low power, low parts count transponders.

To power this hardware - use either the IOIO method - through the USB port

http://ytai-mer.blogspot.co.uk/2011/04/meet-ioio-io-for-android.html

Or more innovatively, using a low baudrate audio modem technique - where both bidirectional data and power is provide through the headphone/mic jack.

http://www.sparkfun.com/products/10331

One earphone channel has full volume 22KHz on it which is rectified and used to power the external hardware, the mic channel and the other earphone channel are used to implement a low baudrate V23 FSK audio modem.

This method - or similar  is used by any device that just plugs into the audio jack - such as the Square, credit card reader https://squareup.com/square


Hope there are some useful snippets there



Ken

[alan hodes]

There are, I'm sure - for the hacker that is interested in giving it a go! Any tips on finding someone? 

[Mark Steward]

Ahh, so you you're looking for a person, not suggestions?

We have a job board for exactly this kind of thing:

  http://wiki.london.hackspace.org.uk/view/JobBoard

Mark

[alan hodes]

That was my original intention - I thought it was clear in my original post (apologies if you didn't think so). But I do enjoy talking about the technical aspects, anyway. Thanks for the tip

[alan hodes]

I've added a new posting to the Hackspace JobBoard - thanks again

On Thu, May 10, 2012 at 2:19 PM, Mark Steward <marks...@gmail.com> wrote:

[Penguin]

Synchronicity:

See Wovyn! and tōd (pronounced "toad") on Kickstarter.com

[alan hodes]

Really interesting stuff, but way over the top for the requirements of this project - as we require only a direct link between smartphone and "sensor" (or rfid fob) and the data doesn't need to be transmitted to a cloud-based database or processed anywhere else other than in the phone itself.