HackRF based SmartNet Radio Monitor

[molay]

Luke Berndt has done a great job with his HackRF recording the entire Washington DC's Fire Dept radio system and making it playable on the web.

http://www.openmhz.com/

Using a HackRF to Capture an Entire Radio System — Luke Berndt

 

[lukekb]

Any Feature Requests?

Thanks for posting! Let me know if anyone has feature requests or questions. I have a couple of ideas on things to add, but it would be great to hear if anyone has ideas.

- Luke

 

[AZScanner]

Thanks for posting! Let me know if anyone has feature requests or questions. I have a couple of ideas on things to add, but it would be great to hear if anyone has ideas.

- Luke

This is really awesome stuff. If I could request one thing, it'd be a detailed step by step on how to set this up for Linux newbies like me (I'm currently archiving off all my files from this computer so I can install Ubuntu on this machine and try this out). I really think solutions like this are the future of scanning and I want to set one of these up for the Phoenix Area. Starting out it will just use one of my RTL-SDR dongles, but if I can get that working I will take it to the next step and use a HackRF and stream it to the web as you have done for DC Fire.

In your blog post you mentioned you were running this on a laptop. What are the specs and what would be the minimum requirements for this? Also, who provides the hosting?

Again, awesome stuff. We need more setups like this out there, and I'm more than willing to hook up the Phoenix area with one of these internet scanners. Just need a little help with the setup, if you're willing.

-AZ

 

[SRSP2282]

Thanks for posting! Let me know if anyone has feature requests or questions. I have a couple of ideas on things to add, but it would be great to hear if anyone has ideas.

- Luke

Wow, very cool work Luke! I really like the ability to filter by talkgroup and tags. Is your software able to handle the decoding of Radio IDs? It would be cool to be able to create a database of interesting/important radio IDs and be able to search for transmissions from specific radios. I'm curious to know what else you are thinking of implementing.

Again, awesome stuff. We need more setups like this out there...

Here are a couple of other sites Luke mentions in his writeup...

Super Trunking Scanner
Radiocapture.com

These are all awesome projects. Guess its time to start saving up for a HackRF SDR

 

[AZScanner]

Working on at least getting GNURadio and DSD to be friends so far... it's been a rough road. After much trial and error today I discovered two things.

1. GNURadio and all the add ons for RTL-SDR and such is a pain to install manually. Do it the easy way: open a terminal window and enter this command: wget http://www.sbrac.org/files/build-gnuradio && chmod a+x ./build-gnuradio && ./build-gnuradio -o then let the computer sit for approx two hours while the script does all the work. Awesome!

2. Luke's DSD code won't compile on GNURadio 3.7. The command above will load the older 3.6 version on your system which is needed for this to work. Found this out after spending 2 hours installing GNURadio 3.7. Ugh.

I'm reasonably certain I can get this thing up and running with my little $20 RTLSDR. A HackRF will be the way to go for anything public facing however. I checked out the other 2 sites Luke mentioned. Very cool stuff, but I didn't see much on the how-to aspect of the project. Hopefully I can change that by setting up one of my own and making a clear, concise, step-by-step, "even a complete Linux newbie can make it work" set of instructions.

-AZ

 

[mancow]

I can't wait for mine to arrive. I wonder if sdr# supports it?

 

[NE1C4NSC4N]

I can't wait for mine to arrive. I wonder if sdr# supports it?

The current version I have for Windows lists the HackRF in the list of hardware it can utilize. I can't say if it actually works or not, but it is certainly there.

 

[AZScanner]

Got everything installed finally, just in time for the forums to go wonky last night. Finally got back in to post today. Has anyone gotten this working? When I run the block I get the following error:

Traceback (most recent call last):
File "/home/phil/gr-dsd/dsd_grc.py", line 511, in <module>
tb = dsd_grc()
File "/home/phil/gr-dsd/dsd_grc.py", line 295, in __init__
self.fir_filter_xxx_0.declare_sample_delay(0)
AttributeError: 'fir_filter_fff_sptr' object has no attribute 'declare_sample_delay'

About all I know about GNURadio is how to spell it, so if anyone can provide a little help here that would be great.

-AZ

 

[SRSP2282]

I'm reasonably certain I can get this thing up and running with my little $20 RTLSDR.

I thought about trying it with mine but I think there are 2 problems I would face.

1. Bandwidth: My RTL-SDR only gets roughly 2mhz of bandwidth. My system covers a slightly larger spread. Hopeful you have a smaller spectrum to deal with.

2. Tuning stability: I have noticed that while tuned to a 6.25khz P25 Control CH the frequency will drift over time. As a result I need to manually intervene and adjust my frequency correction. (Sometimes several times a day) Although the drift is not great, it does degrade my decode rate in unitrunker. Properly tuned I get 99% steady decode, with the drift over the course of the day it can drop down to the 70s. I suspect this would really mess with DSDs ability to decode.

There is some documentation out there that shows how to replace the low quality 28.8mhz oscillator. 28.8 MHz TCXO Option for USB SDRs (mirror) I'm going to do some more digging on this. It would be great to get rid of the drift.

I can't wait for mine to arrive. I wonder if sdr# supports it?

Looks like they added support in v 1145 https://www.assembla.com/code/sdrsharp/subversion/changesets

Apparently there is also a 3rd party plug-in The Midnight Channel - SDRSharp HackRF Plugin Mod Official Website

 

[AZScanner]

I thought about trying it with mine but I think there are 2 problems I would face.

1. Bandwidth: My RTL-SDR only gets roughly 2mhz of bandwidth. My system covers a slightly larger spread. Hopeful you have a smaller spectrum to deal with.

We do have one system I could try it on, Simulcast F of the Phoenix RWC system. Much of it lies between 772 and 774 MHz.

2. Tuning stability: I have noticed that while tuned to a 6.25khz P25 Control CH the frequency will drift over time. As a result I need to manually intervene and adjust my frequency correction. (Sometimes several times a day) Although the drift is not great, it does degrade my decode rate in unitrunker. Properly tuned I get 99% steady decode, with the drift over the course of the day it can drop down to the 70s. I suspect this would really mess with DSDs ability to decode.

There is some documentation out there that shows how to replace the low quality 28.8mhz oscillator. 28.8 MHz TCXO Option for USB SDRs (mirror) I'm going to do some more digging on this. It would be great to get rid of the drift.

Yeah I ran into this also when running the dongle on Windoze, which is why when I get things truly up and running to the point where I feel comfortable sharing the results with the world, I'll be using a HackRF instead. I can't really complain too much about the RTL dongle I got for $20. It's a great way to mess around and prototype things.

Would still love to know how to squash the error I found in the GRC block. Any ideas folks??

-AZ

 

[AZScanner]

I see I've once again silenced a great thread. :roll:

Anyhoo, if anyone is interested, I managed to get the DSD block to run using a very simple flowgraph that takes Audio Source (I've plugged in my trusty BC895's discriminator tap into the Mic port on this PC), pipes it to the DSD block and then feeds it to the Audio Sink for output. Doesn't seem to work (I suspect an issue with this machine is to blame, and not the block) as far as outputting any audio, but the DSD block is chugging along with all the terminal output I'm used to seeing it spit out. I'd love to get it actually decoding though. If anyone has gotten this to work as far as an actual decode from an RTL dongle I'd love to see the flowgraph.

Thanks,
-AZ

 

[lukekb]

Wow, very cool work Luke! I really like the ability to filter by talkgroup and tags. Is your software able to handle the decoding of Radio IDs? It would be cool to be able to create a database of interesting/important radio IDs and be able to search for transmissions from specific radios. I'm curious to know what else you are thinking of implementing.

That's a good idea. It would be pretty easy to pull the ID SmartNet control channel, when a channel grant command is issued. That should be the same unit ID that Unitrunker is giving. DSD also produces a SRC ID from the P25 encoded audio, but it would be tougher to get that.

I was going to add some features more targeted at journalist/bloggers. I am going to try and make calls embeddable in Twitter and also make it possible to build a playlist of calls for a particular event.

The HackRF is definitely a cheap way to capture a wide piece of spectrum. There are a couple of others out there too, like BladeRF.

 

[lukekb]

2. Luke's DSD code won't compile on GNURadio 3.7. The command above will load the older 3.6 version on your system which is needed for this to work. Found this out after spending 2 hours installing GNURadio 3.7. Ugh.

I'm reasonably certain I can get this thing up and running with my little $20 RTLSDR. A HackRF will be the way to go for anything public facing however. I checked out the other 2 sites Luke mentioned. Very cool stuff, but I didn't see much on the how-to aspect of the project. Hopefully I can change that by setting up one of my own and making a clear, concise, step-by-step, "even a complete Linux newbie can make it work" set of instructions.

-AZ

Sorry - I need to put that disclaimer in GR-DSD, and update to 3.7. RTLSDR is awesome, but only works for 2MHz or so of spectrum, so won't cover a whole trunked system. It would be easy to use a couple to tie them together. One benefit of the HackRF is that the tuning error is pretty stable, I think the RTLSDRs tend to drift.

Totally game to help you put together a how-to. There are still tons of rough edges in this space. I built a SmartNet Scanner earlier based on the same approach that has a better writeup: https://github.com/robotastic/smartnet-scanner It would be really easy to do this with 2 RTLSDRs, one for the trunking and another for the audio.

 

[lukekb]

If anyone has gotten this to work as far as an actual decode from an RTL dongle I'd love to see the flowgraph.

Give this flowgraph from my DSD Git. I have gotten it working with an RTLSDR. You will have change the Source in the input block. And also futz with all of the knobs and dials to get the bandwidth and fine tuning correct. One of the tunings is actually negative when you would assume it is positive because of the way GR calculates offsets.

https://github.com/robotastic/gr-dsd/blob/master/DSD.grc

 

[mancow]

Good to hear that, thanks for the reply.

The current version I have for Windows lists the HackRF in the list of hardware it can utilize. I can't say if it actually works or not, but it is certainly there.

 

[AZScanner]

Give this flowgraph from my DSD Git. I have gotten it working with an RTLSDR. You will have change the Source in the input block. And also futz with all of the knobs and dials to get the bandwidth and fine tuning correct. One of the tunings is actually negative when you would assume it is positive because of the way GR calculates offsets.

https://github.com/robotastic/gr-dsd/blob/master/DSD.grc

Hi Luke,

Yeah, that's the one I was using when I encountered the error I posted above. Not sure what I need to fix to make it work, so if you have any ideas I'm open to 'em. Here's the error again so you don't have to scroll back to my previous post:

Traceback (most recent call last):
File "/home/phil/gr-dsd/dsd_grc.py", line 511, in <module>
tb = dsd_grc()
File "/home/phil/gr-dsd/dsd_grc.py", line 295, in __init__
self.fir_filter_xxx_0.declare_sample_delay(0)
AttributeError: 'fir_filter_fff_sptr' object has no attribute 'declare_sample_delay'

My machine may or may not be able to work for this project. The regular DSD runs but doesn't decode anything. I suspect that DSD just isn't gonna fly with Ubuntu 13.10 without some major work.

Thanks for any help on this.
-AZ

 

[lukekb]

Traceback (most recent call last):
File "/home/phil/gr-dsd/dsd_grc.py", line 511, in <module>
tb = dsd_grc()
File "/home/phil/gr-dsd/dsd_grc.py", line 295, in __init__
self.fir_filter_xxx_0.declare_sample_delay(0)
AttributeError: 'fir_filter_fff_sptr' object has no attribute 'declare_sample_delay'

My machine may or may not be able to work for this project. The regular DSD runs but doesn't decode anything. I suspect that DSD just isn't gonna fly with Ubuntu 13.10 without some major work.

Hmm - That is a pretty weird error. Any chance 3.7 didn't get fully uninstalled? FIR_Filters are a pretty basic part of GnuRadio, so it is weird it is causing it to barf. I checked my file and 'declare_sample_delay' isn't in it. That means it is some sort of internal variable. Maybe check you Python Path? Could it be finding something old or funky before it gets to the GnuRadio stuff? Make sure the gnuradio stuff is in your python path. Do simple GRC work in GnuRadio Companion?

Ubuntu 13.10 shouldn't be a problem, that is what I am running. Weird to hear DSD isn't decoding. I wonder if it is some audio driver problems. I think if you run Sax or something like that it makes linux audio more compatabile.

- Luke

 

[AZScanner]

I don't know what the deal is with this machine. Yesterday Ubuntu starting reporting weird errors and then the whole computer decided to crash. I ended up wiping everything and going back to Windows. I'll try this again when I have a machine I can dedicate to running Linux. I don't think this little laptop can do it.Hopefully by next month I can give this another go on a real computer.

-AZ