gambit
PROFESSIONAL
![](/styles/default/xenforo/ranks/General.jpg)
- Joined
- Apr 28, 2009
- Messages
- 28,569
- Reaction score
- 148
- Country
- Location
Now you treads into my area of experience...A long time ago...In a galaxy far, far away...I totally agree with you but here remains a question of prerecorded data. for example the new radars have data bases that have recorded data on aircraft. like there are F-35 flying in Japan and Chinese Radars are recording their RADAR data once a data base has been established the data then can be used to get a better lock with RADAR validating data with it's previous records and thus getting a lock on a stealth fighter. With previous data the algorithm can have a better chance of keeping the lock.
If you look at a model (of anything), the first determination is if the target is symmetrical or not. A sphere is symmetrical. In fact, the sphere is THE model for radar calibration.
http://www.centurymetalspinning.com/radar-calibration-spheres/
https://en.wikipedia.org/wiki/Lincoln_Calibration_Sphere_1
The more simple your model, the less the need for you to view the model from different angles.
When your model is complex like a human being or a car or an aircraft, the minimum quantity of angles is six. The typical front/rear/top/bottom/left/right views. With the human brain, our cognitive capability enables us to recognize targets with less than six, more like four, and with the human face, just one. If the human face turns a few degrees off, we can still call that face 'Joe' or 'Jane'.
The radar's needs are much greater. The ideal situation is to put the target into a controlled environment like a radar anechoic chamber, like this one...The best in the world...
https://en.wikipedia.org/wiki/Benefield_Anechoic_Facility
Then run the radar all over the target. The views you have will be far more than six. The amount of data will be enormous.
But if this ideal situation is not possible, then you resort to the next best thing: morphing algorithm.
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4208213/
http://adsabs.harvard.edu/abs/2002SPIE.4792...78T
Automated target morphing applied to objects in cluttered backgrounds
Basically, the math uses baseline target data, the minimum six angles views, then 'guess' what the target may look like if the view is changed slightly, such as 45 or 44.5 or 87.333 deg off. The variations are practically infinite. Your operating freq is highly affecting to the accuracy of these guesses. The higher the operating freq, the better the ability of the morphing algorithm to accurately guess what the target may look like at 29.593 deg, for example. Granularity -- is the word.
So has anyone managed to get the F-117, F-22, F-35, and the B-2 under controlled environments ? Only US done it.
The tactical implications are staggering for our adversaries. Without revealing anything 'classified', I have said it many times on this forum that the US have effectively defeated 'stealth'.
If you look at the current crop of 'stealth' attempts from Russia and China, you will see very similar planforms and shaping techniques to American 'stealth' aircrafts. That means their 'stealth' platforms will exhibit similar radar characteristics to ours, and we have the most detailed radar views of all.
Whatever radar data the Chinese claimed to have based upon claimed detection of the F-22, assume the benefit of being real for now, those data are junk. And I say that kindly. Simply put, the data did not came from controlled experiments, so whatever morphing algorithm they have, the old programming adage applies: GIGO -- Garbage In, Garbage Out.
That does not mean the Chinese cannot use their own 'stealth' fighters to use as baseline models to try to guess what the F-22 and F-35 may look like. But the US have been flying and testing our 'stealth' platforms for DECADES. The Russians? Forget them. By the time the Chinese are ready, the J-20 and others are DOA: Dead On Arrival.
We will -- not merely can -- weed the J-20 out of clutter and put an AIM-12X up its @$$.