What does router beamforming mean? I tested both the power on and off for a week; my balcony WiFi is all thanks to this

The

principle is most straightforward when shouting slogans. Four loudspeakers shouted in four directions, but none could hear clearly; They stood in a row, aligned with the beat, and shouted to the same person; that person heard loudly, while the others nearby couldn't hear clearly. This is how routers use multiple antennas: transmit the same data, but each antenna is intentionally slightly offset in phase. The signal in your phone's direction is 'in phase' and gets stronger, while in other directions it cancels out and weakens. The total amount of energy hadn't changed, but it had gathered into a single bundle.

the ceiling of beamforming is clear: it saves scenarios where the direction is right and the signal is thinned, not when there are three load-bearing walls and energy simply can't pass through. Wall attenuation is a physical law. I in my on antenna gain calculations, I said the wall consumes the total amount. If you expect beams to form and pierce walls alone, you can give up.

What

Explicit and Implicit Technologies Are: The Technology That Nearly Died Out in the WiFi 4 Era

researching materials, I discovered a little-known fact: beamforming was already included as a standard in the WiFi 4 (802.11n) era, but almost no one actually used it. The reason is that explicit beamforming requires the phone to 'return'—the router sends a probe frame (this frame is called NDP, visible by air packet capture). The phone must measure its own channel status and send it back so the router can calculate how many phases each antenna should be off. The problem is that private implementations from various manufacturers are competing, and mobile support is fragmented and unrecognizable. In the WiFi 4 era, this feature was basically just for show.

WiFi 5 (802.11ac), the standard unified the detection and return processes (explicit compressed feedback). Routers and phones finally spoke the same way, and beamforming truly became a reality. So the current situation is: of invisible beamforming, phones can coordinate with the reply, and the beam is accurate. Phones after WiFi 5 basically support it; implicit beamforming , phones don't respond, routers 'guess' your situation by reading the channel when receiving your data packets. Older devices are also compatible, but their guessing accuracy is a bit lacking.

By the way, MU-MIMO (routers simultaneously sending data to multiple devices in a directional way) operates on the detection mechanism of beamforming—first identifying where each device is, then you can talk about beaming in multiple directions simultaneously. I covered this when I wrote about OFDMA and MU-MIMO, two articles look at each other clearly.

my test of turning it on and off: the real benefits are at the very edge

realized that turning it off was a hidden loss, I decided to seriously test it: I placed the RT-AX86U in the living room TV cabinet, took three points for the whole house, and on the same Xiaomi 13, I averaged five rounds each on and off.

living room sofa, three meters from the router without obstruction, both on and off are above 930 Mbps, so you can't tell the difference—so close, the signal is already overflowing, so whether it's gathered or not doesn't matter. At the head of the master bed, separated by a wall, with 210 Mbps and 190 Mbps open, the gap starts to show but isn't extreme. The biggest change is on the balcony, with an additional wall added over the kitchen. In the very corner of the house: when the switch is off, the signal is -74dBm and the rate is around 150Mbps; when on, the signal is -70dBm and the rate is 210 to 230Mbps. 4 dB signal difference trades for a 60MB speed . Whether my wife binge-watches shows or not is only a 60MB difference. Three sets of data are more intuitive when placed in tables:

this result matches the classic test by SmallNetBuilder abroad: beamforming benefits are concentrated at the signal edges, with almost no visible effect in the central area. The reason is simple—where the signal is strong, the speed has already been maxed out, leaving no room for improvement; In areas with weak signals, the previously wasted energy is picked up, allowing the spatial flow to operate. I calculated the relationship between spatial flow and velocity in the I wrote about 2x2 and 4x4 : when the signal is poor, higher-order modulation simply can't run, and beamforming essentially smooths the space flow path.

free lunch comes at a cost: detection overhead and outdated equipment falling out

beamforming is not free. Every time the router tries to beam, it first has to "shout a shout"—sending a detection frame, the phone measures the channel, and then sends back and forth. This back-and-forth takes up idle time. When there are three or five devices, you feel nothing at all; With twenty or thirty devices packed into the house, the detection frame itself became a new expense. It's the same principle as queuing with many people. I drew a picture of queues in OFDMA .

second pitfall is "channel aging." You strolled around the house with your phone in hand, the beam pointing to where you stood two seconds ago. The router had to re-detect before catching up. Therefore, beamforming provides the most practical gain for the scenario of "sitting still scrolling on your phone," but it discounts the effect for scrolling while walking.

the third pitfall I've really stepped on: old WiFi 4 devices turned against me. My old camera disconnects every few days after enabling 5G manifest beamforming. When I checked the logs, I realized it didn't recognize detection frames at all. The solution isn't to shut down FamilyMart, but to move these old devices to the dedicated 2.4G IoT network, isolating them from the mainline, so new phones can continue to benefit from beamforming. Just like the channel approach, old equipment has its own way of keeping old equipment. How to choose a channel I've written a channel interference special .

Where the switch is and whether to use it: a table is assigned accordingly

different brands hide this switch differently. In ASUS's "Wireless Network—Professional Settings," it's called "Dominant Beamforming." 2.4G and 5G are two independent switches, both on by default; Some TP-LINK models are placed in the wireless advanced settings and are called "beamforming," neighboring OFDMA and multi-user MIMO; Xiaomi and Redmi firmware do not have independent switches; the system has built-in default switches—Xiaomi has an article on their official website specifically about explicit and implicit beamforming. If interested, you can search for KA-167868. No matter how many people search the backend, you can't find this switch. Don't worry, it means the firmware is soldered and it's on by default. If you want to turn it off, there's nowhere to do it. That's actually a good thing.

to wrap up, here's a clear order: the correct way to troubleshoot weak signal rooms is to start with the easy and move to the hardest: first check the signal number on your phone, use anything below -70, and only when it's below -75 do you need to tweak it; Then move the router position. This step yields over a dozen dB of gain, bigger than any switch; Next, check if the channel is crowded; Even after doing all this, it's still not enough—go back and check if the beamforming switch has been turned off by manual labor; After all the fussing and still laging, that's when it's time to upgrade or switch to Mesh. How to read signal numbers and how to position equipment in detail in my on signal strength tuning. To sum it up in one sentence: Beamforming is a good thing, but it just uses the energy you've already paid for, not to generate new energy.