Constructive interference
When two wave sources are in phase, their peaks add. Out of phase, they cancel. A phased array uses this on a grand scale: by delaying each antenna's signal by a precisely calculated amount, you make the peaks line up in the direction you want — and cancel everywhere else.
The phase formula
For an array spaced d apart at wavelength λ, the phase shift needed to steer by angle θ is Δφ = 2π · (d / λ) · sin(θ). Modern chips compute this for thousands of elements every microsecond.
AESA — active electronically scanned array
Each element has its own transmit/receive module. If one fails, the array still works at slightly degraded gain — graceful degradation that mechanical radars can't match. F-35's APG-81 has about 1,200 such modules.
Multi-beam operation
A phased array can form several beams at once: search one volume, track three targets, jam a fourth — all in parallel. This is the killer feature that pushed every modern fighter, AWACS and missile-defence system to phased arrays.