Hello Everyone,
I need certain guidance with my project. I am in need to implement a LoRa(Long range network) on arduino, which is the easy part. For this purpose i have chosen Dragino shield with LoRa protocol as my transmitter/Reciever pair and Arduino M0 as my processor. But what i also need to implement is a direction finding functionality, possibly in the range of 50-100 metres.
While i am aware of certain methodologies such as Watson watt or Ad cock for DF, but all of them come with bulky hardware, while i am to implement this as a much smaller scale factor(Possibly handheld or little bulkier than that).
I had the idea of working with the RSSi parameter, but it doesn't necessarily give the bearing.
Has anyone previously worked with something similar?
Or has any anyone created a miniature antenna array to work with Arduino?
Online, i have found Direction finding projects which are with SDRs and larger antenna arrays. Nothing to work with smaller embedded systems such as Arduino.
Any advice would be appreciated.
No controller can process RF signals in a way that allows to determine a direction. That's why (possibly bulky) antennas are required, which are sufficiently direction sensitive.
What are the objects whose direction you want to find out?
The objects are constantly moving objects. May be drones for example, who with a help of a onboard module are constantly transmitting RF signals to the gateway which is a reciever, which must determine the direction.
What if i somehow made directive sensitive antenna array after all, can Arduino do the signal processing to obtain the direction?
badritripathy92:
What if i somehow made directive sensitive antenna array after all, can Arduino do the signal processing to obtain the direction?
I'd think that you could have a simple sketch to rotate the array with a servo or a stepper. As far as I know (may easily be wrong) the input to an S-meter is just voltage so you could analogRead() the signal strength and have your sketch home in on the maximum reading.
badritripathy92:
The objects are constantly moving objects. May be drones for example, who with a help of a onboard module are constantly transmitting RF signals to the gateway which is a reciever, which must determine the direction.
What if i somehow made directive sensitive antenna array after all, can Arduino do the signal processing to obtain the direction?
But as was revealed here;
There appear to be multiple transmitters ?
Do you think it would be a good idea to give us a full descrption of what it is you are trying to do ?
The Adcock antenna does not need to be rotated, but does need accurate phase detection at either the received frequency or some lower IF frequency.
It's a RF problem and an arduino won't help much.
Check out HF/DF as used in WWII for finding eg submarines.
Allan
srnet:
But as was revealed here;There appear to be multiple transmitters ?
Do you think it would be a good idea to give us a full descrption of what it is you are trying to do ?
You are right. let me write in detail.
For my project, i am designing modules(Arduino plus dragino LoRa hat) which are to be planted on constanltly moving objects,which are transmitting signals based on LoRa Protocol. The gateway must be implemented with the functionality of direction finding these Modules(Find the relative direction of where these transmitters are located ).
And i must direction find for both cases,one where there is one transmitter and second where there are two transmitters.
So while i believe i can direction find it for a single transmitter, Multiple transmitters seems to be a difficult case to implement. Like what if two transmitters are there in the same relative direction.
The constraints are, i need to use Arduino or similar processors and not DSP or SDR, as i need to make the Reciever/Gateway as compact as possible.
You are right. let me write in detail.
For my project, i am designing modules(Arduino plus dragino LoRa hat) which are to be planted on constanltly moving objects,which are transmitting signals based on LoRa Protocol. The gateway must be implemented with the functionality of direction finding these Modules(Find the relative direction of where these transmitters are located ).
And i must direction find for both cases,one where there is one transmitter and second where there are two transmitters.
So while i believe i can direction find it for a single transmitter, Multiple transmitters seems to be a difficult case to implement. Like what if two transmitters are there in the same relative direction.
The constraints are, i need to use Arduino or similar processors and not DSP or SDR, as i need to make the Reciever/Gateway as compact as possible.
badritripathy92:
The constraints are, i need to use Arduino or similar processors and not DSP or SDR, as i need to make the Reciever/Gateway as compact as possible.
Then you will fail.
allanhurst:
The Adcock antenna does not need to be rotated, but does need accurate phase detection at either the received frequency or some lower IF frequency.It's a RF problem and an arduino won't help much.
Check out HF/DF as used in WWII for find eg submarines.
Allan
I will go through them, thank you. I thought of implementing super resolution algorithms such MUSIC algorithm, but i am not entirely sure if arduino processors can handle Eigen matrix manipulation.
aarg:
Then you will fail.
One can always try. May be i can opt for signal processing in SDR later, but i must at least test with Arduino of the limitations or possibilities.
badritripathy92:
One can always try. May be i can opt for signal processing in SDR later, but i must at least test with Arduino of the limitations or possibilities.
How do you propose to overcome the limitations that have been clearly outlined to you in this thread? What signals do your current receivers provide that would allow you to do this?
aarg:
How do you propose to overcome the limitations that have been clearly outlined to you in this thread? What signals do your current receivers provide that would allow you to do this?
As i previously mentioned RSSi parameter of the module can to a certain degree help me to DF, but i believe this can be only for single transmitter. But multiple transmitters, is another case. The processor must be able to resolute incoming RF signals, perhaps I can recieve different RSS indicators and then segregate them based on LoRa device signatures and then find the maximum strength in two set of RSSi values.
But surely, this is just hypothesis and i haven't even started with the programming or the experimental setup.
Any more constructive advice or criticism is welcome.
You can't find direction from signal strength. RF reflects off the multiple objects in the environment and that adds a major random component to the signal strength. You are indeed hypothesizing, and you could save yourself a lot of time by doing some experimentation and/or online research instead of grabbing at straws in a forum.
In the case of a perfect inverse square law for signal strength, you can locate a position (one transmitter!) in 2D with three receivers. That is the best you can do. But in the real world, that won't fly.
I realize that you think having your ideas knocked down is not constructive. But you are wasting your time. If I can save you that, I believe I'm helping.
And no one has asked what type antenna you will be using for receiving. The more directional an antenna is, Yagi, for example, the more side lobes the antenna pattern has. Pick your antenna and then look at the pattern of reception.
Paul
i haven't even started with the programming or the experimental setup
You should do so at the very next opportunity. You will find the experimental approach highly enlightening!
badritripathy92:
One can always try. May be i can opt for signal processing in SDR later, but i must at least test with Arduino of the limitations or possibilities.
Of course you can try.
The use of RSSI to calculate distance (or direction) crops up often on here. Lots of people who have not tried it have this 'idea'.
If there was a workable solution, by now it ought to be well known and documented in a Zillion instructables ................
aarg:
You can't find direction from signal strength. RF reflects off the multiple objects in the environment and that adds a major random component to the signal strength. You are indeed hypothesizing, and you could save yourself a lot of time by doing some experimentation and/or online research instead of grabbing at straws in a forum.In the case of a perfect inverse square law for signal strength, you can locate a position (one transmitter!) in 2D with three receivers. That is the best you can do. But in the real world, that won't fly.
I realize that you think having your ideas knocked down is not constructive. But you are wasting your time. If I can save you that, I believe I'm helping.
Thank you for your reply. Actually, RSSi idea was roughly inspired by this project.
http://fpvlab.com/forums/showthread.php?8331-Antenna-Tracker-Arduino-powered-no-GPS-needed!
Ofcourse one other option that i have is that i can install a GPS module on the transmitter, and then probably could transmit the co ordinates information in the Payload. That makes DF easier.
But i want to see if alternatives exist, so far it seems it either really difficult or non doable. But i am trying to rely more on technology implemented and the data that LoRa protocol transceiver relays than arduino. From my side a more in depth study about understanding LoRa protocol is need, but it seems i might end up using DSPs after all.
And of course not, i dont mind my ideas being knocked down, that is what constructive criticism is all about.
Paul_KD7HB:
And no one has asked what type antenna you will be using for receiving. The more directional an antenna is, Yagi, for example, the more side lobes the antenna pattern has. Pick your antenna and then look at the pattern of reception.Paul
Thank you Paul, I had almost skipped looking at it from the antenna perspective. I will definitely keep that in mind, once the hardware arrives.
Badri
badritripathy92:
Thank you for your reply. Actually, RSSi idea was roughly inspired by this project.
http://fpvlab.com/forums/showthread.php?8331-Antenna-Tracker-Arduino-powered-no-GPS-needed!
Yes, but think about what the project is actually doing.
For FPV you want the receiving antenna pointed in the direction that gives maximum signal strength.
The direction that is maximum signal may not be the direction that points at (as in a map location) the transmitter, in some locations.
For an FPV antenna tracker that does not matter, for finding the actual direction of the object that is the transmitter it matters a great deal.
badritripathy92:
From my side a more in depth study about understanding LoRa protocol is need
Unless your going to use LoRa for transmitting GPS co-ordinates, at which it is rather good, I dont see how a better understanding of LoRa would help.
The LoRa devices can report the signal strength\SNR of the last received packet, so if the packets are tagged with the transmitter ID you can work out a crude triangulation from two seperate receivers, but you will still get the variation in RSSI due to reflections etc.