How WiFi Beamforming Actually Works — Sounding & Steering

Explicit beamforming, standardised in IEEE 802.11ac and carried forward into 802.11ax (Wi-Fi 6) and 802.11be (Wi-Fi 7), requires the client to actively participate by measuring the channel from received Null Data Packets and returning a Compressed Beamforming Feedback Matrix containing CSI data. The access point then computes a precise steering matrix via SVD. Implicit beamforming, used in some 802.11n implementations, infers the downlink channel characteristics from uplink transmissions without client feedback. This relies on channel reciprocity — the assumption that the uplink and downlink channels are identical — which calibration differences between transmit and receive chains degrade in practice. As a result, implicit beamforming produces less accurate steering weights. All modern enterprise access points deploying 802.11ax use explicit beamforming exclusively for its measured superiority in steering precision.

Explicit beamforming typically delivers a 3 to 6 dB gain at moderate distances from the access point. In RF engineering terms, a 3 dB gain doubles the effective signal power at the client, while 6 dB quadruples it. This translates to range extension and throughput improvements at the cell edge where clients would otherwise negotiate lower MCS rates. The gain is most pronounced at moderate-to-far distances; clients very close to the AP already receive strong signals and benefit less. In Abu Dhabi's concrete-heavy commercial environments, where wall attenuation at 5 GHz can exceed 12 dB per partition, the 3 to 6 dB beamforming gain can be the difference between a usable and unusable connection in adjacent rooms.

The sounding exchange begins when the beamformer (access point) transmits an NDP Announcement frame alerting the beamformee (client) to prepare for channel measurement, followed immediately by the Null Data Packet itself — a frame containing known training fields across all spatial streams but carrying no data payload. The client measures how each training field arrived at its antennas, capturing the full channel matrix, then compresses this information into phi and psi angle pairs using a codebook defined in the 802.11ac/ax specification. The client returns this compressed feedback in a Beamforming Report Action frame. Sounding frequency is vendor-configurable, commonly ranging from 100 milliseconds to several seconds. More frequent sounding maintains steering accuracy in dynamic environments with moving clients or reflectors, but consumes proportionally more airtime.

MU-MIMO (Multi-User Multiple-Input Multiple-Output) is fundamentally dependent on beamforming to function. When an access point transmits to multiple clients simultaneously on the same frequency and time slot, it must steer spatially separated beams toward each client while simultaneously computing null-steering weights to suppress the energy each beam deposits at the other clients' positions. This requires current, per-client steering matrices derived from recent sounding feedback. The access point groups clients with favourable spatial separation — ideally located at different angles relative to the antenna array — and applies a composite precoding matrix computed from their individual channel matrices. If clients are co-located or the steering matrices are stale, inter-stream interference degrades all concurrent transmissions. This is why sounding interval tuning and antenna element count are directly correlated with MU-MIMO effectiveness, as discussed in our beamforming and MU-MIMO technology guide.

Construction practices across Abu Dhabi and the broader UAE rely extensively on reinforced concrete with steel rebar, dense masonry block infill, and frequently double-glazed low-emissivity glass — all materials that impose substantial RF attenuation, particularly at 5 GHz and 6 GHz frequencies where modern WiFi operates for capacity. Wall losses in commercial fit-outs commonly range from roughly 8 to 15 dB per concrete partition, with elevator shafts and stairwells creating near-complete RF shadows. In this propagation environment, the 3 to 6 dB gain from explicit beamforming is not a marginal improvement but a material factor in whether a client maintains a usable data rate. Our network design and coverage optimisation process accounts for these UAE-specific propagation factors during the predictive survey phase.

Contact Emirtech Technology at +971 55 651 7843 or submit a detailed brief through our quotation request page at emirtech.ae/request-quotation.html. Our engineering team will conduct a preliminary consultation to understand your premises, client density, application requirements, and construction type, followed by a predictive RF survey and beamforming-capable hardware specification. We serve all areas within Abu Dhabi Emirate including Al Ain, operating from our facility at M44, Mussafah Industrial Area.