First the usual warning – I am both ignorant and intellectually arrogant.  It should be assumed that I don’t know what I am talking about.

Stealth technology hides airplanes, drones, and cruise missiles from radar which uses high frequency radio waves.  It does not hide them from radar which uses radio waves of wavelength a meter or more – the thickness of the stealthy covering would have to be similar to that wavelength.  This means that an F-35 would show up on radar used during World War II.

It has been argued that this doesn’t matter, because radar does not show the location of an airplane unless the diameter of the radar dish is many times the wavelength so an otherwise standard but anti-stealth radar would need a dish 90 feet across.

The key words in the sentence above are “otherwise standard”.  With current technology, it is not technically true that anti-stealth radar would require a dish 90 feet across, as there are phased array radars which do not have dishes (as introduced and used during World War II).  Radar dishes were developed before computers at a time when information processing was strictly analog.  Existing phased array radars are huge devices with identical evenly spaced antennas.  This makes the problem of deciding the phase of the alternating current in each antenna needed to create the desired radar beam simple (it means that phased array radars were used during world war II).  The problem of adjusting the phase of hundreds or thousands of antennas scattered over a landscape (and not all at the same altitude) in order that the radio waves they emit would have an interference pattern corresponding to a desired narrow beam was too hard for WW II information processing. I think it would take this irritatingly slow laptop a millisecond.

I do not think it is difficult to program a large (much larger than 90 feet) phased array of antennas.   By not difficult, I mean I think it would take a computer which costs hundreds of dollars less than a millisecond to do it.  I have a proposal

First scatter antennas.  They can be carried on trucks with engines, dynamos, electronics, a long pole or wire, and, why not a computer more powerful than any in the world when stealth technology was first developed (the additional cost, weight, and electricity consumption would all be negligible). I think it might be best for one truck to carry many antennas with generators and large gas tanks and leave them around (to be guarded and occasionally refueled by soldiers who stay a safe distance away from them when not refueling).

Then it is necessary for the system to figure out where the antennas are (relative to an arbitrarily chosen antenna 1 and in the directions from antenna 1 towards arbitrary antenna 2 and towards arbitrary antenna 3.  For this any frequency can be used (the antennas are not stealthy).  It should take less than a second.  This also means that newly activated antennas can be logged into the system and it can adjust if, say, half of the antennas are suddenly destroyed). 

Then choose phases to create the desired hotspot of electromagnetic radiation. Here instructions have to be sent wirelessly. This shouldn’t take another millisecond. Then send out signals. Then detect returning signals. This takes less than a millisecond.  Then figure out the location and velocity of flying objects in the beam.  This should take much less than a millisecond. All of this works with  meter long (300 megahertz) radio waves.

Imortantly the beam can be shifted as much as one wants from pulse to pulse (so every millisecond).

The time from activation to detection of a stealthy target should take around 100 milliseconds, an interval of time also known as a blink of an eye.

Consider the following vaguely related technology:  multiple antennas connected in a network which can locate working together even though no one antenna can locate by itself.  I call it the “cell phone network”.  The phone company knows where your phone is because it knows the distances from the towers with which it is interacting and has computers which can solve for your location.  Does this approach make stealth technology obsolete ?

Reminded the reader of my total ignorance, I think it does.  There are a number of problems with using multiple antennas in a network to locate an F-35.

First one would have to make the beam head up away from the ground to avoid drowning the desired signal in reflections from hills buildings trees and such. This does not seem to me to be a problem.  Also the targets are fast moving, so the return signal will be Doppler shifted.

The Doppler problem is a frequency pass filter problem which is has been solved for about a century and such that modern electronic equipment can do amazing things. Notably weather radar [update:thank you commenters for pointing out a typo] detects falling raindrops which are much slower than F-35s.  The Doppler shifts detected by different antennas will imply a velocity – each detects motions only towards or away from the antenna but speeds in three directions imply a velocity.

It is also possible to detect a moving object by measuring how the signal changes from pulse to pulse of radar.

Each antenna will not give information on the direction from the antenna to the target. However, the times of arrival of the echo to the antennas will show the differences in distances to the antennas.  Three distances or three differences in distances (calculated from 4 distances plus a constant) determine a location.  The problem of solving for the location of the target (to within about one wavelength) won’t take another full millisecond.

The finding each other, synchronizing, sharing instructions, and all that is done with relatively high frequencies (long enough to get over hills and around trees but not the radar frequencies). Also relatively weak signals (not looking for an echo just looking for the signal itself).

The radiation is strong at points which are a fixed number of wavelengths from dozens of nearby antenna’s, or if the signals are emitted out of phase then an even number of wavelengths from this one and an even number plus a half from that etc.  With many antennas it is possible to focus on a volume which is always a multiple of the wavelength and which is smaller the larger the array and larger the further the target is from the array. The array can be very large.

It seems to me that stealth technology works fine unless the adversary has a digital network.

Importantly it is possible to change the pattern of radiation very quickly changing the frequency, amplitude,  and phase of transmissions from different antennas.  Even if the first echo gives only a very approximate location, it is possible to solve for a much more precise location to fit the pattern of how the echoes strength changes as  function of the phase a  magnitude of transmission from different antennas. This is an information processing problem which is currently trivial.  It is easy to jiggle the radar beam (as much as one pleases every pulse)

Now I think it is necessary for the antennas and processors to communicate wirelessly. This means that it is very important that wireless communications not be jammed or intercepted. This is, to put it very mildly, a very key issue for the US military which uses GPS rather a lot. I think this is another completely solved problem (solved many ways, the cell phone network way, the Wifi way, the HTTPS way, the TV remote control way) importantly one does not have to choose a solution (well TV remote controls are irrelevant) but can use all available anti-jamming and anti-interception tools for one signal.

This also means it is not possible to jam or deceive radar electronically (this is very well known).

This technology is called very long baseline interferometry which is a long established technology which can be used to determine very very very precise images using long wavelengths.  Synchronizing antennas so they work like a huge huge radar dish is another long solved problem.

One shocking aspect of the shocking sinking of the Moscva guided missile cruiser is that the 45 year old ship’s radar could track a limited number of targets at a time and could detect incoming missiles from port or from starboard but not both at the same time.  It is an aspect of traditional radar with a dish that it looks in one direction at a time.

This is absolutely not an issue for multiple synchronized antennas working together. Keeping track of one target or two or two thousand is an information processing problem, which means it is no longer a problem.

I strongly suspect that 100 anntennas and a PC could track 100 F-35s at the same time.  This simply does not seem to be a hard problem to me.  Needless to say, as an economist who knows almost exactly nothing about electrical engineering,  I can’t solve the problem. But I guess that many electrical engineers have solved it just to amuse themselves, while taking a break from working on some incredibly advanced avionics for the F-35.

OK so my 100 antennas and a PC know the location and velocity of a bunch of F-35s. What could one do with that information ? Part of the reason that high frequency radar is used is that the radar system is mounted on an airplane or missile and must be very small.  This is again an information transmission problem. A ground based system can transmit the locations  and velocity of stealthy targets to airplanes and missiles.  The issue is again whether signals can be jammed.  I am old enough to remember a time when it was assumed that missile guidance systems had to be self contained.  I am old enough to remember when GPS guided bombs seemed strange.  I think that this is a currently solved problem which will remain solved roughly at least unless and until quantum computers are developed.

Finally, as I have mentioned before, I think terminal guidance could rely on very much higher frequencies than even modern radar – for example visible light or fairly near infrared. In fact many (most ?) anti aircraft missiles are heat seeking.  However, the antenna system can easily keep track of the location and velocity of a guided missile which is not stealthy and can even contain a transponder.  This makes the guidance problem a bullets hitting bullets problem which is currently trivial.  I guess a guided missile can loop around and try again if it doesn’t hit the target on the first pass.

Importantly, an airplane with a pilot can not accelerate at more than 9 gs. The hardware might manage more, but the human wetware shuts down at more than 9 gs.  An anti-aircraft missile, cruise missile, or (other kind of) drone with a silicon brain can accelerate much faster.  This is one reason why piloted aircraft are obsolete (there are also the problems with oxygen, air pressure, carrying around a body which is just handling controls and all that).

I think the F35 is obsolete.  I am quite confident that the B2 and B21 (expected budget around $200,000,000,000) are obsolete.