18 dbi patch antenna coverage

[edl94]

Hello,
I am new to working with RF and antennas and would like to bounce an idea off of the group. I am developing a system for large gantry cranes that requires I identify the equipment number of the tractor underneath it. I tried RFID but the distance is too great. These cranes work in fairly close proximity 1 meter of separation. I would like to use 2.4 ghz radio modems with highly directional antennas. The crane cab is 43m directly above the tractor, and the tractor is 10m off the centerline of the cab. The crane is 28m wide. Using an 18dbi patch antenna with a 17 degree pattern should cover an area of 21m. My question is will I be able to identify and communicate with the tractor below the cab and not pick up the tractor under the next crane over?
Thank you

 

[VEBill]

It seems like you're asking if you could use such a directional antenna to pick-out a single signal. Probably not very well.

Normal off-the-shelf radio modems combined with 18dBi gain antennas would have pretty long range (~km). Signals would be bouncing all over the place. So you would have to add variable attenuators so that the range was just barely as required to pick out just one signal. And hope the systems were sufficiently shielded that signals don't leak past the attenuator.

But even then you would probably see so much variation over time, and from one to another, that sometimes it wouldn't work at all and other times it might pick up several signals. Balancing unit-to-unit would be a nightmare. In other words, do the math. And then do the math AGAIN but carry the +/- dB variables & error bounds all the way through and you see that it would probably be a 90% solution (at best).

In general, I think that any form of RFID is probably a poor choice for a container port over these sorts of distances. I keep hearing people in that industry talking about RFID, but it seems to me to be a red herring. But I could be proved wrong by a system that actually works.

Technologies that seem more suitable are as follows:
1) Differential GPS with two-way radio link.
2) Vision systems with simple numbers on the roof.
3) Big bar codes.
4) Big QR codes (a type of 2-D bar code).
5) Coded IR emitters on the target with a cheap optics system to pick-out the specific target area you want. The technology is just a souped up 'TV remote control' using the newest high power IR LEDs. Dead easy for trucks with 12vdc DC power. For containers (if required) use lithium batteries and a 2-way IR system to trigger the power-miser emitter. Almost trivial, but I'd add an ACCEPT CODE trigger button and rely on the operator to double check the number visually (add voice output to keep his eyes on the job).

I might have just invented #5 system (probably not the first). The rest are pretty obvious.

 

[VEBill]

Rereading the above, I realize that I sort-of skipped over a primary point.

Any comm system should have a signal margin. In other words, a bit more signal than you need to close the link. Designing the system, you might want +6 dB or +10 dB signal margin. Even the most marginal of systems will still be designed to have a margin.

If you add this into the mix, you'll see that it makes discrimination by subtle amplitude differences an unreliable method.

Then think about reflections (in a container port!) and antenna side-lobes (don't forget those), and the whole thing becomes very unlikely to be feasible. Probably.

 

[Comcokid]

Radio modules using 802.15.4 or Zigbee are being used for asset tracking at much longer ranges than RFID. I have seen demos where several modules were both communicating data and being tracked as the module were activitely moved around. The data was presented on a PC screen hooked to a module with a graphical "map" representation of the different modules locations. Accuracy is typically under 2 meters.

 

[VEBill]

As far as I can see, 802.15.4 and Zigbee are just communications standards. This leaves open the question of what location (or positioning) technology is actually being used?

Triangulation based on RF time-of-flight or relative timing?
Simply carrying GPS data?
Or something else?

 

[MacGyverS2000]

Maybe the suggested "positioning" is simply "Yep, it's still on the premises." K-Marts and Wal-Marts would work this way... items are scanned into inventory as they are unloaded from the truck, then exit inventory when they are rang up at the register.



Sidenote: Will we always call them cash registers once we stop using cash? I'm sure most simply use the term register now, so maybe the entire term will fall by the wayside with the intro of the next generation of shoppers. Hmmmm...

Dan - Owner

http://www.Hi-TecDesigns.com

 

[Comcokid]

Part of the 802.15.4 spec (on which Zigbee is based) specifies that the devices have an accurate RSSI readable in 1 db steps. That combined with TDOA (Time Difference or Arrival), TOF (Time of Flight) techniques between two or more nodes allows one to trangulate the location of devices. The RF range on the 802.15.4 or Zigbee based devices (without amplification) are on the order of a few hundred meters. Companies are already providing devices to locate pallets in a warehouse or cargo containers on a container ship using 802.15.4 devices. edl94 described a situation where the objects to be located are in a limited area - probably a relative straightforward application.

The RSSI, TDOA, and TOF methods for location determination in a limited area are not just for 802.15.4 You can buy WiFi routers that will estimate the location of the wireless devices it is communicating with. Some IT people have used this as an additional security level to make sure those using a wireless link are not outside the business.

 

[VEBill]

The RSSI value obviously isn't going to add much value (except maybe, sometimes, aid the system in guessing when there are two solutions to the equations). A 1dB step-size is a pretty large chunk, and the unknown variables would swamp that out in a complex signal environment.

So basically, it's a time-based add-on and almost completely outside the scope of those comm specs. As I suspected.


"...WiFi routers that will estimate the location of the wireless devices it is communicating with."

Location? or just distance? Distance is easy. Location also requires another distance, or a bearing.


Sorry to nit-pick.

 

[edl94]

Thank you for all of your ideas. You have given me a couple of new avenues to peruse for this project.
Thank you again
Ed