Real time passive radar at home
Vložit
- čas přidán 25. 04. 2024
- Using the blah2 software available at: github.com/30hours/blah2
Using DAB (digital radio) as the transmitter, with a center frequency of 204.64 MHz and a bandwidth of 1.56 MHz. Running a CPI time of 500 ms and processing each CPI in around 600 ms.
See the radar live online at:
radar4.30hours.dev
Antenna geometry is 50% science, 50% alchemy and 100% mystery to me.
You forgot the magic.
@Kansoganix alchemy is just molecular scale aritificery, and artificery is just magic with machines.
I worked for Nokia phones division in 2008-2010. The antenna guys were considered semi-gods who were doing some kind of black magic
You should be able to do it mathematically using Maxwell's equations. Mind you i don't see many people do that.
@@Andrew-rc3vh “Dammit, Jim, I’m a veterinarian, not a mathematician!”
Always love the villian arc from "the government is watching me" to "I'm watching the government"
Whom? says that is a villian arch. you've clearly made a decision about what kind of people do this kind of thing so?? Why are you hear? Its kinda the equivelent of bumping into some one at a porn shop isnt it.
@@sarahmanalapan8443 chill dude, it's a joke not an attack.
Also fyi it's who not whom: who said to whom, he said to him, she said to her, they said to them
@@sarahmanalapan8443 Take it easy, it's a joke.
My comprehension of this is low but still appreciate how well thought out and how well designed this project is. 30 hours sounds like a short time to design this!
Definitely took more than 30 hours! Cheers
That's wild that you can tell it's a bladed aircraft!
Look up “Klein Heidelberg”. World War II passive, static radar system. German…
Props make a dull buzz and turbines make a rocket kind of sound at a distance
Many things have some kind of microDoppler. For example, with a jet engine, if the intake or outlet are in your general direction you can often get JEM, Jet Engine Modulation. When you look closely at the JEM you can potentially see the number of compressor stages and their ratios of operation.
@@FirstToken That's so _cool_
Millitary radar both fixed and airborne can look at the sounds from aircraft jet engines to determine type of aircraft...
Bistatic passive radar is probably the most advanced radar set you can legally own. Well done. This is how radar was kind of invented with civil radio transmitters in the 30s if i remember correctly about Watson Watt's first experimental setup.
I would like to have one too :)
Can't you own any type of passive sdr? So long as you are not using restricted data but open source literature you can probably run whatever you want on receive only?
I think that the effect was first noticed by two radio operators in the US, they noticed that every time a large freighter passed by their noticed it on their oscilloscope. I don’t remember when and where exactly. So if it’s passive radar how is it collecting the signal, is the signal coming from the active radars used at the airport and you are just picking that up. Pretty freaking cool if you ask me.
And even the legality of it is a bi iffy, since several projects on GitHub got pulled by the US government for violating the export laws. :)
@@markoa6999 It sounds like the transmitting power here is coming from a local radio broadcast.
If you have a ham license, there are bands where radar experimentation is legal and effective, BUT: minimum required power rules are still in effect, AND you can't keep it up and running (transmitting at least) all the time.
As a grocery store manager I am glad to learn this
POL56 is a Pilatus PC-12/47 with registration VH-HIG. So not a helicopter but a single-engine turboprop.
You’re correct. The police helicopter here typically takes a POL call sign and I hadn’t seen this one before. There is a subtle micro Doppler difference between a helicopter and turboprop - typically I’ll see more than 1 sideband on either side of the body return for a helicopter.
Even says so in the video
@@30hoursdev You may be aware of this but things like the doppler sidebands from spinning props or turbines have long been used as part of the way military radars can attempt to determine aircraft type automatically. The F-16 has a system called NCTR (non-cooperative target recognition) which tries to give a friend/foe indication based on this information and a database of aircraft types.
it seems the algorithm has just pushed me into rader runners. nice work
Before opening the video: who tf is this?
1 minute in: okay this is some good shit, I'm subscribing
Holy moly mate. Huge amount of work, amazingly done. I checked out the live site just now.
That eyebolt mounting setup is brilliant. You deserve a 10 from the judges for that mate!
I am glad there are smart people like you who can makeup for people like me.
I love how every time you show a new part, you give a reason for the design choice
That coax directly coupled to the driven element of the yagi is reaaaally nice.
Congratulations.
That is one very neat and clean install.
Man, nice organising of the things. Subscribed, looking forward for new videos.
what an awesome way to take advantage of a radio that is transmitting for a totally different purpose. very cool!
Very impressive work. I love how neat and orderly everything is
I've noticed police just don't use ADS-B in Australia. This is an astonishingly unprofessional behaviour, worse perhaps even than their refusal to use normal radio insisting instead every constable be given a digital encrypted radio suite at taxpayer expense. I suspect these luxuries are because they do not want public oversight, which begs the question; what're they hiding? We all know the usual excuses they will have, but the bottom line is we pay for their service, all records are public, transmit in the clear and turn your transponder on ffs. Even the MILITARY are blasting on ADS-B just fine. But the police in Australia truly think they're the SAS or something. They sincerely need a speaking to by your citizenry and a reminder of whom they exist to serve.
That’s all Aussie cops for you. They think they’re above the rest of us, rather than serve us.
Thanks for sharing your experiments and results. This is a very detailed video and the system was very clearly explained.
That is pretty neat! Tahnk you for sharing, looking forward to more videos!
Cool. I didn't know there was such a thing as passive radar (and initially I was going to question this video. But some quick reading before posting saved me from the embarrassment).
Very informative!
I wish i could be this smart! Thanks for the video😊
Very nice setup.
Reminds me a lot of the things we do in radio astronomy.
Really nice and clean setup. Love it!
This is most definitely not my area, but I love listening to subject matter experts talk, without Bob & Jane’ing the material. Outstanding. Sub.
Excellent leading-edge work! All the best, Rob in Switzerland
Very cool. I didn't know this could be done. Thank you.
This is SICKKKKKK need one of those round displays where your in the center.
Incredible work!
Very sharp mate! 🌈
Instant sub, this is awesome work.
New sub this randomly popped up in my recommendations (I'm a Ham) and holy crap instant sub to this channel . Well done and I will share this with our club members to check out
Hope to have some content on active radar (transmitting in the 70cm television allocation) and fox hunting - VK5DST
Your construction and organization of all the components is beautiful I have a comment or two on the things that I saw in your video.
The yagi antenna shown is somewhere in the 200 MHz region, using eye bolts to hold the elements is genius, love it.
At frequencies above one gig coax of any length is a bad idea, ALL R.F. components Must stay on the antenna at the focal point.
You will learn to hate grouping your RF connectors in a tight space, type 'N' are precise making a connection and then getting them tight requires a perfect straight in alignment, very difficult to do if they are clustered together.
Awesome content love what you've created impressed with the outcome thanks for sharing.
dE AA7J
That's an awesome setup!
...best part of your setup is the squeaky mouse-wheel. 🥰
incredible work
Superb work.
This was awesome. I hope to see more videos on the future. I agree with you previous commenter about being able to detect which aircraft were bladed. I'd love to know how you're able to differentiate between bladed & non.
Planes have the doppler of the fuselage, wings, etc. On top of that is the doppler of the parts that are moving. Each one has a signature depending on the number of blades, speed, size. Jet engine blades are fast but can be hard to see because they're ducted so you won't see them at most angles. Props are easily visible. Helicopter blades huge relative to the size of the aircraft and generate doppler in the plane of your radar.
Very cool! I cant get over how much like Norton Sparkles this guy sounds tho lol.
very interesting setup
Amazing stuff!
This was super cool!
Passive is craziest part about this. This is insane
This is awesome!
Fab would be proud!
What a world we live in!
This is so cool!
Wow 1 video and 1.8K subs already!
Damn, that's a cool project!
now that's some proper shenanigans
This is really spectacular for something done DIY
I have also experimented passive radar, I am looking ways where a hybrid radar system could be built, using Ham bands to use three modes, Passive, line of site and over the horizon, on a common frequency ranges. As I have done a lot work with radio navigation and have a new type of signal processor design, with few improvements could be used for Radar. As there so much technology out there now, where it possible for anyone to play around with as a hobby, to get an idea how radar systems work. Good to see what you have done so far.
Transmitters are expensive; I spent 47 years designing and building them at the AF Research Lab. Using somebody else's transmitter is way cheaper but you have to compensate with processing.
Hams don't transmit enough power, and they're not on long enough to be reliable. Broadcast FM radio and TV are strong signals on high towers that you're surrounded by.
Line of sight: yeah, that's pretty much the limitation. Although if you look at high flyers you'll catch signals off them from beyond line of sight. That's a problem for getting a reference signal. You'll also get high power microwave pulse radars as they swing by.
OTH active HF radar that reflects off the ionosphere takes much, much more power than you can afford, to detect targets. But you can play with passive detection from meteor burst and stuff. We played with all that kind of stuff at the AFRL radar test range in Verona, NY. There's science to it, but a whole lot of luck when you talk about propagation around the atmosphere.
@@jamesvandamme7786 There is a lot to it, at the moment I have been playing around with phased antenna arrays with the Class Q RF amplifier design. That has I & Q inputs, where each antenna has it own power amp, by controlling the phase on antenna / amplifier block will change the beam of the array. The limitations are max RF power levels you need work within on the ham bands, is a major issue for sure. It is a good time do experiments with radar as there is so many different ways to try out new ideas. As it is all good fun to build and test some of these ideas out.
Impressive !
That's some mindblowing stuff right there, i have no idea what 99% of this equipment does or how it functions but it's amazing nonetheless.
wow - thats amazing !!
Awesome project! Regarding that coax on the dipole, don't you need a balun for that?
That coil of coax on the boom acts as a choke balun.
Thats so cool!
Impressive 😮
Impressive.
I woke up today with my mind set on trying to make a 2.4Ghz microwave radar. A few hours into researching some bits and bobs and I've ran across this fantastic video!
My thinking is make a Patch Antenna and direct it at a parabolic dish lined with aluminium foil. I might need to get a custom circuit printed as I feel a breadboard setup would be very innacurate and have too much inherent parasitic inductance and be very hard to get the impedance accurate enough. But I want to get a signal being sent out (Probably about 0.2 Watts?) and just get the data recieved into my computer where I can go crazy processing it (I have a Data Engineering background, this should be the easiest bit ahahah)
Then set it up on a gimbal and leave it for a while to scan in my room, and see if I can process it into an image. Probably using an arduino for stepper motor control and just sending my computer the position data with each read.
Do you have any advice, or see anything difficulties/errors in my thinking :)
Afterwards I'd like to extend it to work with Pulse Doppler and try tinkering around to remove some of the static.
All the best.
That’s the problem with going higher in frequency, you can’t get away with the breadboard desktop setup that you might at lower frequencies. I’ll point you towards the MIT coffee can “cantenna” radar which gives you a list of parts you can put together for a higher degree of success. I understand the key part of the radar processing is done in hardware with a mixer, rather than the SDR approach here which does all the steps in software.
It would be nice if you could explain what the terminology you used means and how the components actually function.
Nice video! What brand are the lightning discharge tubes?
Wonderful setup! Can I ask which gas discharge tubes you used and how did you select them? I assume because the antennas are passive and only receiving that there is very little voltage across the GDT under normal conditions.
Selected the cheapest ones available on AliExpress. The purpose of these is under lightning conditions, to discharge any static charge through to the ground stake. I haven’t actually connected the body of the connectors/antenna interface box to the ground stake yet, but should do it soon. Otherwise these devices should have a minimal insertion loss.
Super cool
How neat!
Cool project. Might you want to swap out that ferrous hardware in the antenna?
this rocks!
What i font get is the polarisation of the reference is digital radio not vertically polarised where you live ? Or does the polarisation not matter for the surveillance antenna and only the reference antenna should be oriented correctly?
The reference is vertically polarised. The surveillance is h-pol to avoid interference on the radiation pattern from the vertical mast. This also reduces the amount of v-pol clutter (which is the dominant signal source by far). It’s hard to quantify without doing a full simulation in the environment of the receiver, but target SNR is strong on the cross-pol. A good experiment would be to run a dual-polarised Yagi and directly compare target SNR.
So 04:18 and 04:32 the X axis is the distance to source (simply). And the Z axis (vertical) is the direction from transmitter and receiver.
The thing needs a proper graphical user interface. A one window where the primary is the "radar" scope.
B Scope for range and azimuth as secondary in vertical form, but as well primary should be way to get a circular polar coordinates from your position.
That you would get simply circle where your antenna is at the center, and you get azimuth (0-360 degree) and range (in km), as in "233 degree, 5.5 km." And a way to select a direction that is for calibrated heading, so you could example use a radio to give you North.
And have a way to enter coordinates for your antenna, and system would calculate a GPS coordinates from that to each detection.
Later on have a detection memory, like a submarines, where each contact is automatically assigned a name. They can be the radio brevity alphabet for starts to have just those 20+ contacts listed in time order. The challenge would be to include a tracking capability, so if contact is lost for moment, then use the prediction of existing area (circling aircraft) and vector and speed that if contact appears there, it is likely the same previously known target instead a new contact.
X axis is loosely distance, Y axis is Doppler (positive coming towards, negative going away) and the Z axis is the target strength.
This 2 channel system does not have any concept of angle of arrival. All I know is the target is at some range on an ellipse surrounding the transmitter and receiver. Target localisation can be achieved using an antenna array and beamforming, or a multistatic approach like in the 3lips project.
Agreed that a tracker would provide redundancy for any missed detections.
I suddenly have a strong desire to make my own radar setup for at home
damn ive been looking for this type of video for couple of weeks and lost interest... and now it got recommended to me... well i hope i'll learn something
This is absolutely amazing that you can do such things at home these days. Any chance you might do a homebuild ground search radar at some point?
At 2 MHz bandwidth in this system, each range bin is around 150 m. If you design a ground penetrating radar to see a few meters into the ground, an appropriate range resolution can be set by increasing the transmitted bandwidth. I would expect a bandwidth of hundreds of MHz would be required, which would be out of scope for this system. Interesting idea.
@@30hoursdev First off thanks for responding so quickly. Secondly I had actually meant more along the lines of the sort of ground search radar used nowadays in both vehicles and in military units. To detect other vehicles, people, other objects etc. Or I guess the best known civilian use would be radar cops lol. The ground penetrating radar is definitely something I hadn't considered but I agree it is a most interesting idea....
@@AGTheOSHAViolationsCounter I see! Unfortunately the same principles apply of needing a lot more bandwidth. Radars on vehicles are generally in the tens of GHz (out of reach for cheap SDR platforms) and require high bandwidth if you want resolutions in the tens of centimetres or lower. My understanding is a lot of the signal processing on these vehicle radars is done in hardware.
Great work ! I had no idea an sdrplay could be used as such a granular reference for looking at microdoppler. if so, one could review this method on recorded IQ data. Are you using their API?
The project does have the ability to record IQ data to file for post processing. A hidden functionality of pressing space bar (but not on the public radar). I use the SDRplay API to grab samples to processing buffers.
I wonder if you could widen the observed spectrum with more receivers at different bands to gather more data and further increase the detection and resolution by overlaying the different data feeds. Imagine having a receiver for every detectable radio tower to get a bunch of feeds and then aligning them all on a few displays to add better range and direction finding.
fascinating, so how exactly is your radar finding flight numbers? is it listening in for the ATC signal coming from the aircraft? or is it just trying to correlate the radar returns with a flight control map?
and all the signal near the 0 shift line, is that corellated to the transmitters? or is it a reflection of the skyline?
also will you make any project to improve the computer here so it can track contacts specifically? ie discerning individual targets, finding its speed/heading/altitude, and displaying it as a clear target?
Listening for ADS-B signals transmitted from the aircraft, to correlate with the radar returns. The returns at 0 Doppler are static clutter, which have been removed using the clutter filter. My other project 3lips on GitHub uses multiple radar nodes to geolocate the target on a map.
@@30hoursdev ah OK, thanks for the info!
I wonder if there is a way to view the radar on a Toyota? they are 3mm band and seem to have good range on them.
Interesting idea! The software defined radios I use only go up to the low GHz region, where the car radars are in the high 10’s of GHz. Signal processing is more computationally intensive with larger bandwidths.
Thats awesome
super cool project, I took a peek at your other one as well that is even cooler. If you want a hand optimizing your code, I would be happy to help you out but its not clear to me if you are computer starved or not, I poked through your code and have a few ideas but I didn't see your test data file in the repo.
Writing unit/functional tests is lagging and I haven’t added golden data yet. Not really starved computationally but always happy to hear ideas on the Discord.
Very cool
So cool
Awesome project !
At 0:37, the length of unbraided coax is way too long, and is making part of the dipole length. So i doubt the antenna is well tuned, you could probably improve it A LOT by making this connection as short as possible. The yagi elements support solution is smart !
Agreed! I should include the feed into my NEC model and quantify that. With regards to matching I modelled a beta/hairpin match but never installed it on the antenna. I measured a VSWR of 1.75 and since the antenna is receive only, it’s just a (relatively small) attenuation loss.
@@30hoursdev True, but keeping in mind that VSWR is not always a proof of good efficiency, specially with all the coax you are using, that can somewhat hide or attenuate the mismatch.
Also feeding a yagi with a dipole is often not optimal, but with the type of mount you are using, you could easily switch to a gamma match, with possible good improvement :)
Thanks for sharing your project ! I have always found passive radar super interesting .
Impressive
Been waiting for this one. I have a SDRplay RSPdx, can I use that?
It will need to be the SDRplay RSPDuo as it’s the only model with the dual tuner. The reference and surveillance channels need to be synchronised.
The craft you called a helicopter is a turboprop fixed wing.
You’re correct. The police helicopter here typically takes a POL call sign and I hadn’t seen this one before. There is a subtle micro Doppler difference between a helicopter and turboprop - typically I’ll see more than 1 sideband on either side of the body return for a helicopter.
@@30hoursdev I remember when listening to AM on a tube radio (longwave, or shortwave?), when ever a helicopter was starting somewhere nearby, I could hear a hissing distortion quickly growing faster with its increasing rotator speed, and the green sector on the magic eye (signal level display tube) jittered at that speed too. As a kid I mused if the magic eye could be modded into an actual radar screen to display the position of those helicopters.
thats awsome
Wait... passive radar? I want to know more! Any chance of doing more videos on the fundamentals of this?
Easy answer: same as regular radar, but someone/somewhere else is transmitting the signal.
@@d10valentin perhaps I should have been more specific.
I gathered as much as you said from the word "passive". I'm more interested in the implementation, the fundamentals as in the building blocks of a solution, not the fundamentals as a layperson's explanation.
@@lmaoroflcopter You should find fundamentals inside software. A lest all match operations.
Look for projects like Celldar from the 80s and 90s that tried using reflected signals from phone masts. One massive problem was aerial design on phone towers stopped it working as they steered radiation downwards!
It worked for vehicles but was never good enough and was abandoned, so don't waste your time and money on a project uk government abandoned.
Awesome! Do you have a ground rod near the antenna interface box?
I do. Not shown but a meter away from the antenna interface box is the house ground rod. I still haven’t run a cable connecting the body of the box to ground, but will soon.
Great work! Could you tell the part number/ search term for those gas discharge tubes? they look really neat
Search online for “SMA lightning arrester”, I purchase mine off AliExpress. You can also purchase the internal gas tubes separately at places like element14, Mouser, etc.
@@30hoursdev awesome! big thanks
what is illuminating the targets? I recall from a while ago that researchers used local radio and TV stations for the illumination
Digital radio signals (DAB) with 50 kW power.
6:10 "the clutter filter and the ambiguity processing are the most demanding parts of the system"
Welcome to my brain...
Impressive work and setup, whats the max detection distance you've plotted?
Difficult to generalise since it’s dependent on target RCS, geometry and the waveform being used. Also note bistatic range is not true target range from the receiver. The system I have running with DAB as a transmitter is on the order of 20 km true range.
@@30hoursdev Still amazing and impressive!
Not bad at all.. reminds me of radar sets that would have been used in WW2.. so can you array this (like 16 receive antenna)?
For sure! The SDRplay RSPDuo only has 2 channels, but I could use something like the KrakenSDR with 5 channels, or various USRP devices. Will have to re-architect some of the code to support phased arrays.
Subbed 😊
This is a really interesting idea.
It is a nearly perfect UFO detector. Most aircraft have transponder signals. Any that don't would not be identified and hence would be a UFO.
I subscribed now.
I am sure there is another use this idea can be put to but I can't think of one right now.
UFO would have stealth technology
Some of those UFOs will be aircraft that are purposely excluded from the data flow to civilian consumer radar display apps.
Assumes UFO was not spoofing someone else's ADSB code ...
@@uploadJ That would be a MSO a misidentified flying object.
@@uploadJ Good point! Maybe someone can write code to flag alien spoofing.
WOW!!
Why use coax for 5vdc and not regular cable? Is there a benefit? If so, does the polarity matter?
It was convenient since I already had the coaxial run done. It was neater to keep everything over coax rather than add new connectors to the antenna interface box. Polarity does matter, ground is connected to the cable bulkhead, so you don’t want 5V shorting out there. Keep the 5V on the inner conductor.
Hi, sorry for the really dumb question, but i sont exactly understand how it works.. why the two antennas? Who's the emitter? Could you make a short video explaining more or less how the convept works? Thanks!!
There’s a digital radio transmitter which sends signals all over the city. The reference antenna receives the transmitter signal directly, while the surveillance antenna receives signals that have reflected off aircraft. Then I do some signal processing to generate a delay-Doppler map and show targets.
@@30hoursdev Makes perfect sense; thanks!
@@30hoursdev Subsidiary question: I guess both signals (reference vs surveillance) must be perfectly time-synchronized right ? Because some µs of difference translates into many km of difference.
Correct! The SDR used is the SDRplay RSPDuo which has 2 synchronised channels.
@@30hoursdev So, would that be cancelling out the "noise" or capering data samples to spit out a result?
Norton’s been busy
Mom, can we have radar
No we have radar at home
radar at home: