Stream1090

FlightAware ADS-B Flight Tracking
620

I am getting errors when compiling stream1090.

image
image1020×605 85.4 KB

I have the right version of cmake. I am using g++ 15.2.1. I started the build by making a subdirectory and running cmake ../ in it then I typed make.

I can’t easily figure this out.

1 Like
621

Thank you very much. I messed up in the case that no airspy and no rtlsdr support is there. It is fixed now.

622

Thank you. It now compiles and I get an executable.

1 Like
623

Let me know what you want to pipe in there. IQ Format and speed.

624

I wanted to pipe in 2e6 sampling rate and 16-bit signed integers for the IQ values.

I had to up sample 2e6 to 2.4e6 and cast down to 8-bit integers for the IQ values.

It’s just for a quick test to compare what stream1090 gets in number of messages from a single antenna versus four antennas and different beamformers.

625

are these signed integers using the full 16 bits?

626

Yes, signed integers but not using the full 16 bits. It’s 11-bits and twos complement for the negative numbers.

In my code, I divide by 2049 but I think the correct number is 2048 to get full scale 0.0 to 1.0 floats.

627

Well can you feed floats between -1.0 and 1.0 into stream1090?

628

Yeah, I can feed it floats. I can feed it 32-bit or 64-bit floats.

629

we do float32. Give me a bit of time for custom input 1 IQ pair = 2 floats between -1.0 and 1.0

630

Pull a new version

cmake ../ --fresh -DENABLE_CUSTOM_INPUT=ON && make

with ./stream1090 -h you shoud see something like this:

Stream1090 build 260219_2
(custom input mode)
....
Supported sample rate combinations:
  2  →  2 (float32 IQ)
  4  →  4 (float32 IQ)
More Messages with Beamforming
631

I pulled the newest version, ran cmake, and then make. But the final binary does not show support for the float type. It also gives the following warning message about -DENABLE_CUSTOM_INPUT=ON.

image
image692×559 48.3 KB

632

Sorry, its late i forgot to push the CMakeLists

633

Yep, that fixed it. I might test it in the next 15 minutes or a few hours from now. I can’t tell but I will let you know if it doesn’t work or something looks wrong.

You’re C++ is very good. I like how you set it up.

634

Thx, the current code was a bit of work, but it is flexible while at the same time remains fast. I needed that float32 input mode anyways. I haven’t tested it yet, i will do it tomorrow.

635

One thing you should know:

This will put stream1090 into passthrough (and stupid) mode. For 2 → 2 it will create two bit streams. It will compare x[i] > x[i+1] for i even (first stream) or i odd (second stream). Or in other words:

Stream 1: x[i] > x[i+1]
Stream 2: x[i+1] > x[i+2]

For 4->4 it will create 4 bit streams

Stream 1: x[i] > x[i+2]
Stream 2: x[i+1] > x[i+3]
Stream 3: x[i+2] > x[i+4]
Stream 4: x[i+3] > x[i+5]

IF (<- really IF) your IQ pair streams are captured at the same time. You can interleave two IQ pair streams @2Msps and feed them with a sampling speed of 4 instead. Even then complete duplicate removal is not guaranteed yet.

Edit: If it works and you need more. Well, the beauty of the current solution is that you have:

#if defined(STREAM1090_CUSTOM_INPUT) && STREAM1090_CUSTOM_INPUT
constexpr auto presets = std::make_tuple(
    // Custom Input
    Preset<IQ_FLOAT32, Rate_2_0_Mhz, Rate_2_0_Mhz, IQPipelineOptions::NONE>{},
    Preset<IQ_FLOAT32, Rate_4_0_Mhz, Rate_4_0_Mhz, IQPipelineOptions::NONE>{}
);
#else

and there is

enum SampleRate {
    Rate_1_0_Mhz  =  1000000,
    Rate_2_0_Mhz  =  2000000,
    Rate_2_4_Mhz  =  2400000,
    Rate_2_56_Mhz =  2560000,
    Rate_3_0_Mhz  =  3000000,
    Rate_3_2_Mhz  =  3200000,
    Rate_4_0_Mhz  =  4000000,
    Rate_6_0_Mhz  =  6000000,
    Rate_8_0_Mhz  =  8000000,
    Rate_10_0_Mhz = 10000000,
    Rate_12_0_Mhz = 12000000,
    Rate_16_0_Mhz = 16000000,
    Rate_20_0_Mhz = 20000000,
    Rate_24_0_Mhz = 24000000,
    Rate_40_0_Mhz = 40000000,
    Rate_48_0_Mhz = 48000000
};

For passthrough mode, the compiler will put together what you want without any more changes except in the preset table.

More Messages with Beamforming
636

Conversation moved to the beamforming thread. Here things will focus a bit more on the filter and rtlsdr business. Currently i have a lack of proper 2.56 Msps data. If someone has the patience to capture some for me, that would be really nice.

637

I have a 2.56 Msps file. 2.9 GB, How do I get it to you?

1 Like
638

Can you try: https://www.swisstransfer.com/ and send me a link? It is free and does not need an account.
Edit: Or use some other free site of your choice like https://www.file.io/

639

Link to file samples.raw is: SwissTransfer - Send large files securely and free of charge

Edit: I thought that would just be sent via e-mail and not shown here. Is there a way to do that?

On the other hand, if someone wants to download that large file (2.9 GB), I guess they’re welcome to it. Transfer rates are kinda slow though.

1 Like