Broadcast TWT
Understanding Broadcast TWT
Legacy Wi-Fi power save requires stations to wake at every DTIM beacon interval (typically 100 to 500 ms) to check for buffered frames, consuming significant power even when no data is pending. Individual TWT (802.11ax) improved this by letting each station negotiate a custom wake schedule, but the per-station handshake creates overhead with hundreds of IoT devices. Broadcast TWT solves the scaling problem: the AP advertises one or more group TWT schedules in its beacons, and stations autonomously join a group based on their traffic pattern.
The AP divides associated stations into groups (for example, 8 groups), each assigned a staggered TWT Service Period (SP). Only one group wakes at a time, reducing contention by the number of groups. Within each SP, the AP can use OFDMA trigger-based access to schedule transmissions deterministically, eliminating random contention entirely. Stations sleep between SPs, with intervals ranging from seconds (interactive devices) to hours (environmental sensors), enabling coin-cell battery lifetimes measured in years.
Power and Contention Calculations
D = Tactive / Tinterval
Example: 16 ms active / 300 s interval = 0.005% duty cycle
Power Savings vs Legacy PS:
Savings = 1 − (DTWT / Dlegacy)
TWT: 0.005% vs Legacy: 0.5% ⇒ 99% savings
Contention Reduction (N groups):
CollisionTWT ≈ Collisionlegacy / Ngroups
100 STAs / 8 groups = ~12 STAs per SP
Wi-Fi Power Save Mode Comparison
| Mode | Standard | Wake Trigger | Negotiation | Min Sleep | Best For |
|---|---|---|---|---|---|
| Legacy PS | 802.11 | DTIM beacon | None | 100 to 500 ms | Basic clients |
| U-APSD | 802.11e | Trigger frame | Per-AC | Variable | VoIP, real-time |
| Individual TWT | 802.11ax | Negotiated schedule | Per-STA | Seconds to hours | Latency-sensitive IoT |
| Broadcast TWT | 802.11ax | AP beacon announcement | None (group) | Seconds to hours | Dense IoT sensors |
| Restricted TWT | 802.11ax | AP beacon + enforce | None (group) | Seconds to hours | High-density venues |
Frequently Asked Questions
Broadcast vs Individual TWT?
Individual TWT requires per-station negotiation, creating overhead with hundreds of IoT devices. Broadcast TWT eliminates this: the AP announces group schedules in beacons, stations join autonomously. Broadcast scales to dense deployments; Individual provides tailored wake times for latency-sensitive devices.
How much battery savings?
A sensor waking every 5 minutes for 16 ms uses ~0.005% duty cycle vs ~0.5% for legacy PS (waking every 300 ms DTIM). This 99% power reduction extends coin-cell battery life from weeks to years. Actual savings depend on sleep interval and active duty cycle.
How does it reduce contention?
Stations are divided into temporal groups (e.g., 8 groups). Only one group wakes per service period: 100 stations / 8 groups = ~12 per SP, reducing collision probability from ~30% to under 5%. The AP can further use OFDMA trigger-based access within each window to eliminate contention entirely.