Author Topic: The opposite everyone is trying to achieve... low distance transmit power  (Read 3291 times)

DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Hello all,

My first post, so be gentle.

I'm planning on creating an indoor location system based on motion moteinos that would have a proper antenna length and design and set on high power to communicate as they normally should to a gateway.

What I'm trying to achieve is some necklace FOB with a purposely low transmit range (undetermined at this time, hence the question) that would periodically broadcast on 255 to any motion moteino's in range. Those will have a sketch that will forward which FOB IDs it is seeing at the time and possibly RSSI. Not relevant but they would also act in the traditional PIR function.

So my question is: Has anyone played around with purposefully low range? I'm not talking 30cm on a lab bench but anywhere around 6 to 12 feet or anything around there? Particular combinations of antenna design and RFM69 power settings?

Not looking for someone to bench it for me, just wondering if I an save a bunch of time on research and I can start benchmarking it myself and tweaking.

Let me know if you have any ideas. Thanks!

TomWS

  • Hero Member
  • *****
  • Posts: 1925
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #1 on: December 20, 2016, 08:59:46 PM »
Are you asking if range can be purposefully limited to a 6 to 12 feet range?  If so, this radio is NOT the one for you.  You probably want to look into Bluetooth Low Energy or some such.  The RFM69 receiver has such a huge dynamic range that it would be very hard to limit its range to such narrow confines.

If you're asking if the RFM69 can be used to set up near range ad hoc networks, absolutely.  You just need to know what the limits are. 

Frankly, it sounds as if 'range' isn't the issue as much as 'exclusivity' of traffic, ie, limited to a very select 'audience'/'respondents'.  In this case, the RFM69 really shines because you have control of frequency, 'network id', and encryption, all factored to limit the 'audience' of any traffic.

Let us know what's really important and I'm sure there is someone on this forum who can help you!

Tom


DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #2 on: December 20, 2016, 10:04:31 PM »
Thanks for reaching out.

You are correct, arbitrarily limited based on antenna design and power settings. I've been reading posts on here and elsewhere that others have gotten very low range but not what they intended: Anything from needing to touch the antennas together (not setting the transmit power to high on certain models) to different power settings on the data sheet to bad (or in my case good) purpose built antenna design.

Range *is* the key but does not need to be super precise, just within a number of feet so that the 'low range' beacons (small low consumption key FOBs) that transmit would only reach the normal and properly set up RFM69 nodes within that range.

Use case would be:
1) user (keyFOB) is in range of sensor A = user is in a particular room
2) user is in range of sensor A and B = user is a particular room, or between rooms, or we are just interpolating
3) user was in range of sensor A and/or B and is now seen in sensor C = they moved from one area to the other
*) just a few types of examples and all dependent on resolution (range and placement of sensors) and can be tweaked based on further knowledge

Most indoor location systems require a smartphone as the receiver and low power beacons. My goal is to:
1) not require that you carry a smart phone around your house/office/establishment to have the system work.
2) use an existing RF and protocol that would already be implemented in nodes around the environment to reduce cost
3) those nodes could already be performing other functions (PIR, flood, temp/humid), again, reducing cost
4) not require my keyFOB be charged nightly in the case of BLE (or maybe that is not much of a concern, must look into it).

My only concerns/questions:
1) depending on the range of the keyfob (including it hanging around your neck and you body being a good absorber/reflector of RF energy) how would I have to space my sensors around an environment to get an 'arbitrary' resolution of where someone might be.

Not sure if I'm being more informative or just muddying the waters even more.

TomWS

  • Hero Member
  • *****
  • Posts: 1925
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #3 on: December 21, 2016, 09:36:47 AM »
Given your explanation of the application, I would DEFINITELY recommend using Bluetooth Low Energy.  There are probably off the shelf modules (fobs) that already send out beacons.  All you then need is a BLE Host module at each site to detect the presence of a beacon.  Presumably each site would be able to be Mains powered and therefore, you could tie the Site's BLE module into an ESP8266, which, in turn, could compare its findings with the other sites and then post the result to a local node (Rasp-PI, etc) or the cloud...

EDIT:  I would definitely NOT use an RFM69 for this application.

Tom

DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #4 on: December 21, 2016, 11:41:12 AM »
Thanks TomWS for your input. I was hoping for that non-existent technical holy grail where one size fits all ;)

I'll shift gears and take a look at BLE. I'm aware of iBeacons but they work opposite to my use case. I'll see what I can dig up.

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6371
  • Country: us
    • LowPowerLab
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #5 on: December 21, 2016, 02:44:12 PM »
@DeltaZuluCharlie,

As an afterthought and FWIW take a look at these threads which talk about antenna matching using small antennas including 50ohm resistors.
Originally intended for "bench-range" testing, the idea is that a close to 50ohm impedance match will absorb most power from the radio and dissipate as heat, but may yield the small range you're after.

https://lowpowerlab.com/forum/rf-range-antennas-rfm69-library/inverted-f-antenna/msg15843/#msg15843
https://lowpowerlab.com/forum/rf-range-antennas-rfm69-library/small-loop-antennas-433-mhz/msg16427/#msg16427
https://lowpowerlab.com/forum/rf-range-antennas-rfm69-library/small-loop-antennas-433-mhz/msg16461/#msg16461

Chip antennas also perform "poorly" compared to monopoles/dipoles, something to consider in any application given your requirements.

DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #6 on: December 21, 2016, 02:51:17 PM »
Interesting... typical to most 50ohm loads for other RF systems. Some cool research indeed from others.

Now I have to decide if I want to tinker with that or go BLE where I could also map 'external' users/visitors to the system that have cell phones.

Have to love technical architecture. Cost vs. Time vs. Functionality/Expansion. ;)

Really appreciate the information Felix. Thanks.


joelucid

  • Hero Member
  • *****
  • Posts: 867
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #7 on: December 21, 2016, 04:14:13 PM »
Why do you need to restrict range to detect proximity? Can't you just measure RSSI and measure which Moteino picks up the strongest signal from the fob? If the PIR nodes are already mains powered that would be trivial to achieve and would avoid having to install two devices in every room.

The only downside I see is that you have less attenuation at the lower frequencies. So other factors - like antenna directionality will be a stronger factor in what you measure. But i don't see any fundamental reason why this can't be done with rfm69's.

DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #8 on: December 21, 2016, 04:25:43 PM »
Joe,

The reason is the unpredictability of RSSI based on reflections, absorption. I'm not an RF engineer and my math skills are at best a first year university student level so triangulation is out, heat maps are out... not really but I don't want to over complicate things.

The simplest, most reliable and cost effective way I came up with was battery powered sensors that would already be placed in rooms (perhaps multiple), hallways, etc. that would be performing other functions such as the Moteino. Then use a range restricted keyfob also using a low cost RF solution as a beacon that deep sleeps.

This would have allowed me to keep the same protocols across multiple sensors and the only thing that needs to be reported to central 'business logic' would be, "user A is in range of sensor 1' and the logic only has to figure out that means, "user is in the Kitchen".

I don't need to know exact x,y,z coordinates. With enough cheap sensors that can only receive the beacon within x feet I can figure that out to determine an area/zone the person is in.

At least that was my thoughts. Still playing with ideas.

joelucid

  • Hero Member
  • *****
  • Posts: 867
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #9 on: December 21, 2016, 04:32:49 PM »
Quote
The reason is the unpredictability of RSSI based on reflections, absorption. I'm not an RF engineer and my math skills are at best a first year university student level so triangulation is out, heat maps are out... not really but I don't want to over complicate things.

I think RSSI is very closely related to whether a packet can get received. So if RSSI values are an unreliable predictor then packet reception likely is, too. And I do think 2.4 ghz would yield better proximity measurements for the stated reasons. But I think it would be worth a try.

But if the sensors are battery powered you'll have a hard time to detect the fob since both would have to be awake at the same time. All doable but much less of the shelf so then it might make sense to just use bluetooth as Tom has suggested.

DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #10 on: December 21, 2016, 04:37:47 PM »
Very good point.

Between sleep states and collisions it may be very unpredictable depending on how fast the beacon is moving and how responsive I want the system to be.

Ah, I would be discouraged if I didn't love a challenge.

Thanks for your thoughts.

TomWS

  • Hero Member
  • *****
  • Posts: 1925
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #11 on: December 21, 2016, 06:48:06 PM »
I'm aware of iBeacons but they work opposite to my use case.
I humbly suggest that you look at iBeacons again to see how they work and what their purpose is... (hint: A SmartPhone or Tablet IS a BLE Host with respect to iBeacons)

Tom

WhiteHare

  • Hero Member
  • *****
  • Posts: 1297
  • Country: us
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #12 on: December 22, 2016, 11:10:49 AM »
If you do go the RFM69 route, maybe you could use gobs and gobs of gateways to help locate your fobs (sorta rhymes, doesn't it?).  i.e. if you were to put 4+ gateways in every room and look at received RSSI from your fobs, then by the sheer density of your gateways you could probably narrow down the fob location reasonably well without resorting to triangulation (which might not work so well anyway).

ziplockk

  • Newbie
  • *
  • Posts: 28
  • Country: gb
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #13 on: December 22, 2016, 12:07:01 PM »
Concur with BLE beacons - as used in short distance range finding applications in lots of domains these days ;-) . . .

For example to track people who wear them on a pendant around their necks so that an automated system can detect whether they have stopped moving, or where they are in a building . . . add in a pressure sensor and you can detect whether they have fallen or just stopped moving . . . if they happen to be prone to falling and getting hurt for example . . .


TomWS

  • Hero Member
  • *****
  • Posts: 1925
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #14 on: December 22, 2016, 05:52:00 PM »
if they happen to be prone to falling and getting hurt for example . . .
LOL!  Maybe I should get a couple!

Tom

DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #15 on: December 23, 2016, 01:04:30 PM »
I humbly suggest that you look at iBeacons again to see how they work and what their purpose is... (hint: A SmartPhone or Tablet IS a BLE Host with respect to iBeacons)

Tom

I have indeed looked into it further. It seems the prices have come down tremendously since I first looked at BLE transceivers a couple of years ago along with a lot of DIY/open source community development.

My concern at the time, and up to you suggesting I take another look was that I didn't want the iBeacons on walls and having to carry around a smart phone all the time. I wanted the opposite but without expensive iPhones or whatever on the wall.

Looking at the nRF51822 right now for the room sensors and still poking around for 'fobs'. TI has some interesting sensortags.

DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #16 on: December 23, 2016, 01:08:49 PM »
Concur with BLE beacons - as used in short distance range finding applications in lots of domains these days ;-) . . .

For example to track people who wear them on a pendant around their necks so that an automated system can detect whether they have stopped moving, or where they are in a building . . . add in a pressure sensor and you can detect whether they have fallen or just stopped moving . . . if they happen to be prone to falling and getting hurt for example . . .

Dang you Ziplockk. Now I have to integrate an accelerometer into my fobs. :P

TomWS

  • Hero Member
  • *****
  • Posts: 1925
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #17 on: December 23, 2016, 06:56:57 PM »
I have indeed looked into it further. It seems the prices have come down tremendously since I first looked at BLE transceivers a couple of years ago along with a lot of DIY/open source community development.

My concern at the time, and up to you suggesting I take another look was that I didn't want the iBeacons on walls and having to carry around a smart phone all the time. I wanted the opposite but without expensive iPhones or whatever on the wall.

Looking at the nRF51822 right now for the room sensors and still poking around for 'fobs'. TI has some interesting sensortags.
The model I'm suggesting is to use an off-the-shelf fob/tag, similar to what @ziplockk mentioned, as the key chain/fob/bodypiercing/whatever that gets carried around.  These all send periodic beacons.

At each point where you care about detecting the presence a 'body' carrying such a beacon, you have a 'station' that consists of a BLE host/client/whatever term du jour is being used (BlueGiga has a couple of usable ones) that, in turn, is connected to a gateway device (ESP8266 is a reasonable choice, but RaspPI, or any other programmable device that can bridge between the BLE world and a LAN will work), which revolves the closeness of the 'presence', possibly comparing with other stations, to determine where, exactly the 'body' is with respect to the stations.  The concept is trivial and workable as long as you're prepared to fill in a LOT of gaps in coding/hardware integration.  I'm not aware of any affordable off the shelf solutions.

Tom

WhiteHare

  • Hero Member
  • *****
  • Posts: 1297
  • Country: us
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #18 on: December 24, 2016, 06:14:03 PM »
Perhaps using the "Tile Mate" approach would be easier?  Just make the receiver beep audibly, and then let your ears guide you to it.

Felix

  • Administrator
  • Hero Member
  • *****
  • Posts: 6371
  • Country: us
    • LowPowerLab
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #19 on: January 01, 2017, 04:56:22 PM »
FWIW I ran across this article where the author (good credentials) says:

Quote
Triangulation on BLE is impossible because on 2400 MHz you will have a lot of reflections, multipath plus phase shift.
Further down:
Quote
A detection on room level is possible. If you add Kalman filters and some more sources like the magnetic field of the earth, accelerometers or even ultrasonic sound then the result will be much better than on BLE beacons only.

Maybe worth a read.

DeltaZuluCharlie

  • Newbie
  • *
  • Posts: 9
  • Country: ca
Re: The opposite everyone is trying to achieve... low distance transmit power
« Reply #20 on: January 01, 2017, 04:59:24 PM »
Thanks. I've also been reviewing this one over the holidays: http://www.cl.cam.ac.uk/~rmf25/papers/BLE.pdf