Think of radio transmitters as little stones dropped in a pool. You know from high school physics that a dropped object will send out waves across the water’s surface. If you drop two stones, those waves will overlap with each other in a regular “interference” pattern. Changing the characteristics of a stone will change the amplitude and phase of the waves it emits, as well as the characteristics of the interference pattern generated with waves from other stones.
If you have enough control over the situation, you can have a sensor at the edge of the pool looking for just the right wave pattern, and you can keep changing the stone characteristics until that exact pattern arrives at that particular point. Elsewhere in the pool, the wave pattern will be different, and that’s fine. You’re only looking for that one pattern in that one place. Everything else can be ignored.
In a nutshell, this is the essence of beamforming. You’re controlling the output characteristics of each transmitter within a transmitter array so that the overall signal is optimized to reach a given receiver in a given direction. With an antenna array in which each antenna is transmitting with slightly different characteristics, you have what’s called a phased array. As we’ll see, there are two primary forms of phased array used in wireless access points: on-chip and on-antenna, adopted by Cisco and Ruckus Wireless, respectively.