RADAR and Stealth

This is as simple as they come:

The F-117A (original Stealth Fighter) was said to produce the radar signature of a sparrow.
Impressive hiding job, guys.

But why couldn’t the bad guys crank up their radar and, when telling it what to display:

  1. Anything over 50’ long
  2. Anything with an measurable area over 30 sq ft.
    3, Any sparrow traveling more than 200 mph.

Since everything is digital, real-time screening of radar returns (seems to me) would be instant. So why not look for anything moving fast enough even it it looks like a bumblebee?

Or would a radar array that sensitive become too clogged for even a NSA-sized analysis system to decipher?

Yes, have ben wondering about this since the first Revell kit all those years ago*

    • the pentagon denied the existence of the F-117A until Revell released a credible model of it in 1981

You are fighting signal to noise. This really is as fundamental as it comes. You are asking for information about the subject from the radar return. The spectral signature of the return may tell you a few things. A frequency shift gets you velocity. Change to the pulse profile can get you some information related to the shape. But all of this requires that you can extract the signal out of the noise. You can’t avoid noise in the system. Everything from thermal and shot noise in the radar receiver, ionospheric noise, jamming, general clatter of other systems, thermal noise, all the way down to the cosmic microwave background. There is noise. As the intensity of the return signal from your subject drops your signal to noise drops. That limits the amount of information you can extract from the return signal in a precise manner that is as hard and fast as the laws of thermodynamics. The only way to haul the signal out of the noise is to make lots of measurements over time. Time is the one thing you don’t have. So the information you can glean is reduced to the point that you simply can’t perform the filtering you desire.

Radar doesn’t give you the length and width of an object. If you have ever seen movies where they have a submarine and its sonar sends out a “ping” and then they listen for the echo, radar is essentially the same way except with radio waves instead of sound. The radar sends out a radio “ping” and listens for the echo.

There are two main aspects to stealth. One is to absorb as much of the incoming radio wave as possible so that it doesn’t bounce back, which is where the F-117’s fancy black skin comes from, and the other is to channel all of the radio waves that aren’t absorbed in certain specific directions so they don’t bounce back to the radar receiver, which is where the F-117’s funny shape comes from. As Francis Vaughan said the end result of this is that you get such a weak return that it is difficult to actually track it as a signal. You don’t get a reliable 400 mph sparrow. You just get a very weak signal that is difficult to pick out of the background noise.

The Revell model also had pretty much nothing to do with what the Pentagon did or did not admit to. There was plenty of speculation at the time and Revell just took the most popular ideas and turned them into plastic. I personally wouldn’t even call it all that “credible” of a model either. There was no faceting and it really didn’t look much at all like the real F-117.

An F-117 was shot down over Yugoslavia. I believe they were able to track it by playing around with the frequencies that the radar was using. The F-117 is also a bit easier to track when it opens its bomb bay doors, so that may have factored into it as well.

Case in point:

Revell F-19 model

Accurate F-117 model

Did anyone have digital radar in the early 80’s?

This article makes it sound like practical digital radar is very new if the Navy is not yet done converting old systems. Signal processing time has been a significant hurdle - it’s OK if digital weather radar takes ten seconds to process signals and decide if it’s raining or not, but for tracking fast aircraft, it’s probably pointless to say “There was an airplane over there.”

AFAIK, the technology was still experimental into the late 80’s.

Air Traffic Control radar is a whole different ballgame. Most of the problems have not been anything to do with the digital processing of radar signals, but to do with the mangement of the flight data once received. The insane sorry tale of the FAA ATC upgrade is beyond belief. Billions spent and almost nothing to show. The AAC system is one of the biggest snafus in history. Trouble, is they are at it again. Four years ago it looked bad. Two years later it looked in serious trouble. It still isn’t any better.

ATC systems are not representative of radar in general.

Digital processing of radar has been around for a while. A lot depends upon what you want to do, and how much money you want to spend. Cabinets filled with very expensive custom logic implementing DSP systems were not exactly rare. Even if your mobile phone could outpace them in every metric you can think of.

Man that model looked way cooler than the real thing.

FWIW, I seem to recall that during the Kosovo kerfluffle, one F-117 was downed by a SAM operator who later claimed he’d figured out how to calibrate his radar to pick the plane up - but he also had additional data to work with, namely he’d identified a pattern in the bombings that had occurred in the previous days/weeks that suggested the planes always egressed using the same route and focused his radar dishes there.
IIRC he’d purportedly also found out a particular wavelength against which the F-117s coating & angles didn’t work as well as they did against the wavelengths typically used by missile systems ; and he had to do some electronics work to get his dishes to work on that new frequency.

For its part, the USAF denied it and claimed it was a fluke, one-in-a-million lucky shot.

Here’s the Wiki page on the guy.

I don’t know where that “signature of a sparrow” thing came from.

I do know that I achieved radar lock on an F-117 from the backseat of an F-4. The blip on my screen was the same size as I’d expect from any other fighter, but not quite as bright. Also, I knew exactly where to look, as the plane took off right in front of us.

If you can’t detect an object, because it just isn’t reflecting a strong enough signal to appear above the inherent noise in the radar, then you’re not going to get any information about its speed or size.

So what about the first part? Cranking up the power.

The more powerful a radar you want, the bigger it will be, the less mobile it will be, and the more expensive it will be.

So if you had more powerful radars, you’d have fewer of them, and they’d be in positions that were less of a surprise. So they could just be flown around or attacked with anti-radar missiles.

Also, any extremely powerful non mobile radar is just asking to get destroyed by an Anti-Radiation Missile.

If you’re a SAM commander going against a modern air force, you have 20 seconds to locate and lock onto your target and get a missile on the way, because the 21st second your radar is on is the last second of your life as an AGM-88 HARM get’s sent your way.

Radars wouldn’t be able to detect a sparrow going any speed, at any reasonable distance. The amount of the radar signal an object echoes back is measured in special anechoic chambers, very isolated from outside noise, and up-close. Larger objects are measured outside, but still relatively close, like a few hundred meters.

Adding distance to this makes for a big problem - the amount of signal power echoed by an object and picked up by the receiver goes down by the fourth power of distance - something twice as far away will give 1/16 the signal back at the radar site.

So you could measure a sparrow up close in an anechoic chamber, but not for any distance that you’d want to be able to detect a bomber.

And completely unairworthy; from the blunted canards to the submerged intake ducts on the upper body well aft of the cockpit, the only way that thing could have left the ground was in another aircraft.

The F-117 was1970s-era technology, developed long before real time digital signal processing was possible, and the stealth technology developed for it (both the materials and systems used for design simulation) are obsolete today, which is why the system has been retired.

As far as dialing up the gain or looking for small phase changes, as already indicated, this gets into the point that signal to noise is less than unity, unless you find a particular band at which you get a characterisic response, and then have the capability to band pass filter to isolate that signal. This is an area where techniques developed for radio and x-ray astronomy have actually filtered back to military applications.


Are you sure about that year? I remember the F-19 (what the F-117A was conjectured to be called before the DoD admitted its existence) models being a late 80s/early 90s thing, not early 80s. (Even had the Microprose game F-19 for the C64.)

While hardly a dead-on match, Testor’s MiG-37B model, which I had, ended up being closer to what the American stealth fighter looked like, with its more angular look.

You left out the words “totally awesome” between “the” and “Microprose.”

For a very understandable example of signal to noise:

A few years ago we were using analog televisions. Some cable systems still use this.

If you tune to a channel that has no signal, what you see looks like a snowstorm. This is what noise looks like. The television receiver is turning it’s RF gain all the way up trying to receive a signal that isn’t there. Because the gain is turned all the way up, any stray signal from lightning storms, worn out electric drills, astronomical objects, and** especially** the thermal noise generated within the electronics of the receiving circuits themselves gets scaled up to a full blown white dot. Since these events are random, or at least uncorrelated with the scanning frequency, you get random white dots at random times. So in this blizzard of randomness, you are trying to see the one snowflake that is making steady progress across the screen. Good luck with that.

So it turns out that when you are looking for a very weak signal, you need a receiver that doesn’t generate very much of it’s own noise. This is hard. If the signal is steady, and not moving, then you can keep averaging over time, and the noise will eventually average to zero if it is truly random. If the target is moving then you need advanced, refrigerated electronics to see these very weak signals. If the signals from space and crappy drill motors are much stronger, then you STILL won’t catch the weak signal.

To be fair, you’d probably say something similar of the F117, if you didn’t know it was real.

And Oakminster, I assume that you were above the stealth plane? The shape is probably optimized versus detection from below, since that’s where most enemy radars would be.

Glad to know that I’m not the only one who wasted a summer away playing that game. :slight_smile: I should fire up the ol’ emulator and see if it’s as good as my memories…or maybe I should just leave my memories be.

Ages ago there were reports that the Oz over the horizon backscatter radar systems could pick the steath planes without any issue at all. Again mostly because they lit the plane from above. The wavelengths were a lot longer too. I would imagine the old Soviet OHBR aka “woodpecker” radar probably had a similar capability.

I would suggest the latter. A recent revisit to MicroProse’s F-15 Strike Eagle III only brought disappointment. I rapidly expunged the recent memories and stayed with the glorious old ones. Oddly enough, those were summer too. Ah, yoof.

I was wondering that, too. And, did the F-117 driver kindly ask you to knock it off?

PS- Oakminster was a WSO? That’s exceedingly cool.