A-RTLS (Assisted RTLS) is a common method for RTLS protocols in the Leantegra system, where L-MESH supplies the presence notifications about nearby devices to the RTLS anchors. This mesh-assisted mechanism is used together with Unified TDMA to enable the most optimal usage of batteries and RF traffic.
A-RTLS is not a positioning method. It is a layer over the methods the platform already has — UWB TWR, UWB TDoA, L-DCB and Offline-online RMA — that changes when they run rather than how they measure. The measurements are unchanged; what changes is that the radio producing them can sleep.
An anchor's job is to listen. Left to itself it cannot know whether any device is nearby without listening speculatively, and on the positioning radio that is expensive:
| Radio | Receive current |
|---|---|
| UWB | ≈ 130 mA |
| 2.4 GHz | ≈ 4.6 mA |
| ≈ 28 : 1 |
An anchor that opens its UWB receiver across the ranging region to find out whether anyone is there draws on the order of 115 mA — which is not a duty cycle a battery device can hold. Speculative listening, not measurement, is what makes battery-powered positioning infrastructure hard.
At low occupancy an unassisted anchor spends more energy discovering absence than measuring presence. Where devices are in range 10% of the time, nine wake-ups in ten find nothing, and each still costs a receive window.
A Unified TDMA sync beacon establishes the time base, so every synchronised device knows where the slot boundaries are. It does not, and cannot, say whether anything will transmit in a given slot — and that distinction is what A-RTLS addresses.
The asymmetry is between the two roles. A tag derives its slot from its own identity: it knows its EUI, the schedule maps that to a slot, and sync alone tells it when to wake. An anchor cannot do the same, because its work depends on other devices' identities, and no clock accuracy reveals which devices are within range today.
| Sync beacon | Presence | |
|---|---|---|
| Answers | when does slot N begin? | which devices are in range? |
| Direction | master to everyone, broadcast | device to anchor, announcement |
| Needed by | every device, to participate at all | anchors, to know which slots are worth opening |
| Without it | unsynchronised — no slots, no participation | slots opened speculatively, most find nothing |
Sync is the timetable; presence is the passenger list. Neither substitutes for the other, and a perfectly synchronised anchor still does not know whether anyone is boarding.
Three properties make this work, and all three come from the shared schedule rather than from any new measurement:
| Property | What it gives |
|---|---|
| Presence on the cheap radio | the anchor learns a device is in range for ~1/28 of the energy |
| Deterministic slot mapping | the ranging slot follows from the device identity and the schedule — nothing is exchanged to agree it |
| One time base | a wake time computed on the 2.4 GHz side is valid on the positioning side |
The benefit tracks absence, not presence. Assistance saves the wake-ups that would have found nothing, so its value rises as occupancy falls and reaches zero when devices are always in range:
| Occupancy | Unassisted | Assisted | Gain |
|---|---|---|---|
| 100% — continuously present | — | — | 1.0× |
| 25% | 3.3 months | 5.2 months | 1.6× |
| 10% | 4.6 months | 12 months | 2.6× |
Anchor service life, five devices ranging at 2 Hz. Because the gain is zero at continuous occupancy, presence gating is a runtime setting — an anchor on a continuously busy route disables it and saves the 2.4 GHz cost instead.
Assistance is method-agnostic. What differs is how much each method has to gain, and whether it remains viable at density.
| Method | Accuracy | Assisted by | Note |
|---|---|---|---|
| UWB TWR | centimetres | presence gating | an exchange per device; airtime scales with device count |
| UWB TDoA | centimetres | presence gating | one transmission heard by every anchor; scales further |
| L-DCB | metres | runs on the mesh itself | no separate radio to wake |
| Offline-online RMA | zone | relay of buffered records | positions where there is no coverage at all |
Method choice is bounded by device density as well as by accuracy. A two-way exchange costs airtime per device, so at high densities in one zone the one-way and mesh-based methods remain available where two-way ranging does not — see UWB TWR and UWB TDoA for the airtime arithmetic.
The anchor holds what it has heard. The cache is not bookkeeping — its ageing policy is a power setting, because every entry that outlives the device it describes buys a wake-up that finds nothing.
| Field | Purpose |
|---|---|
| Device identity → slot | derived from the shared schedule |
| Last heard | refreshed by either radio — a successful measurement is also evidence of presence |
| Consecutive misses | silence in an assigned slot is strong evidence of absence, and cheaper evidence than waiting for a timeout |
| Capacity | bounded; a roaming population must not grow it without limit |
Ageing is a trade with no free setting: too aggressive and a lost announcement costs a measurement; too slow and departed devices drain the anchor. A failed measurement should age an entry faster than time alone does, since it is evidence already paid for.
A change to the schedule invalidates every cached slot mapping. The cache is discarded on any schedule change rather than migrated, because a stale mapping wakes the radio at the wrong moment — which looks like a measurement failure rather than a configuration error.
| L-MESH | carries presence, and relays the results |
| Unified TDMA | the shared schedule both radios read |
| A-GNSS | the same assistance principle applied to satellite positioning |
Current and lifetime figures on this page are derived from device datasheet characteristics and the schedule arithmetic — UWB receive at 130 mA, 2.4 GHz receive at 4.6 mA. They describe the design and want confirming by measurement on a bench unit before they are used to size a deployment.