Do I have the following right: Tracking works by making range/angular/velocity gates which act as blinders into which energy is focused. At each cycle, the computer determines the strongest return within the gates and recenters the gates around that point.
Is it accurate to say that a radar in targeting mode is much like a radar in tracking mode, the differences being that targeting usually uses continuous wave rather than pulses, the freqs are higher for higher resolution, the gates are smaller for greater precision and the updates are more frequent? IOW: targeting is tracking on 'roids.
Say 4th generation fighters are using their X-band radar against each other, how large would their range/angular gates be and how frequent would the updates be for search, tracking and targeting modes? I realize there can be significant variation in this, I’m just looking for ballparks and especially the ratios between each mode.
Is LPI radar mainly useful for search rather than tracking and targeting? LPI works by aggregating faint signals over time. This works well for search. However, in tracking mode, you need to keep the target within the gates at each update otherwise you can lose your track, right? That means needing a certain minimum frequency of updates which may not allow aggregating signals over time. Therefore, a radar may have to increase transmitted power above the noise floor, correct? With targeting, the gates are even smaller so the updates must be even more frequent which leaves even less room for LPI. Is that accurate or am I missing something?
It’s probably roughly right. Keep in mind the people who can speak with authority on this subject, can’t. It is a felony for them to reveal secret or top secret details of modern X-band aircraft radars, and all of what you are asking for are going to be secret or top secret. All people can do from an armchair is look at the fundamentals and try to come up with a rough guess of the capabilities.
There are many ways to do the signal processing, and the way you mention is just one of them. It gets a lot more complex for systems that high end, and I only have a vague idea myself.
And. something like low probability of intercept is going to be the most secret of them all - think about it. If you knew how it really worked (you’re a Russian, and an American traitor gave you the real specs and spectral data from LPI systems actually running), you could probably make a reliable detector for it, defeating the whole purpose of it.
You’ve also mixed up several generations of thinking there. One sentence is accurately describing how 1970s radars worked. The next sentence is applying that logic to cutting edge 2017 prototypes. That’s defective.
What did I get wrong? Which sentence is wearing the bellbottoms? The first few questions weren’t just a lead up to the last one, I was curious to know about those too, even if some of the info concerns now-obsolete radars.
The crude way you describe tracking - by simply figuring out where all the energy is reflected from a contact and continually refocusing more and more of your electronically steered beam to the most reflective point. Then simply inferring from the state of the radar and the state of the aircraft it is mounted on the velocity and position of the target contact.
Among other problems, that algorithm would fail catastrophically if a semi-stealthy aircraft were to release some sort of chaff or decoy. Now, the brightest thing in the sky is not the main body of the aircraft, and the radar is going to waste all it’s energy on that. It will stick the box around that bright bit of metal, ignoring the real threat.
The truth is, the real techniques for a 2017 radar aren’t just officially secret - you would realistically need a PhD in the field and 20 years experience to understand all the tricks. I say this as a coder-monkey/engineer who uses tricks like that which I only have a vague, intuitive grasp on myself. There may be ways to genuinely differentiate between stealthy contacts and chaff. There may be prioritization algorithms so the radar always keeps updating anything in the sky that fits a far more complex set of parameters for potential threats. Sensor fusion would in theory let you correlate IR and visual light camera data (if you are close enough) to sort out what should be the primary threat.
And, another very real limitation is all this cool sounding stuff might be on the software roadmap but the F-35s leaving the runway right now may have none of it actually working and just far cruder methods.
It’s crude because I’m only looking for a high-level understanding in order to reproduce it in software form. I’m not asking non-PhDs for a PhD-requiring answer.
I’d have thought the chaff problem would be solved by using Doppler filtering. This should work unless the chaff-launching aircraft is traveling perpendicular-ish to the radar beam.
Well, that’s something I can talk about. I know right how to develop such a model.
What you need is the radar equation and you need to just arbitrarily decide on a couple thresholds, and you can model all other ranges.
You need not bother with any of what you talked about, just assume the aircraft’s systems can see the contact if it’s within the threshold (you might decide that you can see the Su-35 at 300 km with the F-35 radar, head on, and if the Su-35 is viewed side on, that’s about 40% more reflectivity and you feed that into the equation for a detection range. Then as you iterate over the data structure of the battlespace, if the distance between the F-35 and Su-35 is less than the scaled detection range, and the F-35’s radar’s central cone is facing the Su-35, it shows up)
You need not bother with any of the other details. There are too many unknowns you simply don’t have specific information on. (like under what conditions, less than what is on the data sheet, with the F-35 radar be overloaded?)