Ultra-wideband does not estimate distance from signal strength — it measures how long a pulse takes to arrive. Radio travels roughly thirty centimetres per nanosecond, so the whole problem reduces to timing a flight to within a nanosecond or so. Everything difficult about a UWB deployment follows from that sentence.
A narrow pulse in time requires a wide spectrum. Ultra-wideband occupies hundreds of megahertz precisely so its pulses are short enough to time accurately, and short pulses are what separate a direct path from its own reflections.
That is the whole advantage over signal-strength methods. Received power varies with obstruction, orientation, body proximity and reflection at least as much as with distance, so an estimate built on it is noisy in exactly the environments where location matters. Time of flight does not care how strong the signal is, only when it arrived.
The arithmetic is unforgiving. At roughly thirty centimetres per nanosecond, a one-nanosecond timing error is a thirty-centimetre position error. Sub-metre accuracy is therefore a sub-three-nanosecond timing problem, which is why UWB hardware is built around timestamping rather than around radio sensitivity.
| Two-way ranging | Time difference of arrival | |
|---|---|---|
| How it works | Tag and anchor exchange messages; the round trip gives the distance | Tag transmits once; anchors compare arrival times |
| Clock synchronisation | Not required between devices — the round trip cancels offset | Anchors must be tightly synchronised with each other |
| Airtime per fix | Several messages per anchor pair | One transmission, heard by many anchors |
| Tag battery | Higher — the tag transmits and receives | Lower — the tag only transmits |
| Scales to | Modest tag counts with good accuracy | Large tag populations |
| Suits | Access control, tool tracking, precise interaction | Large estates tracking many assets |
Two-way ranging removes the synchronisation problem and pays for it in airtime and battery. Which constraint binds is a deployment question rather than a technical preference.
By timing how long a radio pulse takes to travel, rather than inferring distance from signal strength. Radio covers roughly thirty centimetres per nanosecond, so sub-metre accuracy is a sub-three-nanosecond timing problem.
Two-way ranging exchanges messages between tag and anchor so the round trip cancels clock offset, requiring no synchronisation but costing airtime and tag battery. TDoA has the tag transmit once and anchors compare arrival times, which needs tightly synchronised anchors but scales to far more tags.
Because geometry determines accuracy as much as the radio does. Anchors close to a straight line give good accuracy along that line and poor accuracy across it, so the same hardware produces very different results depending on where it is mounted.
Multipath and blocked direct paths. A reflection can arrive stronger than the direct signal, and when the direct path is blocked the signal travels around the obstruction — which biases the measurement long rather than making it merely noisy. Averaging does not remove a bias.
It depends on update rate, because airtime is finite. A few tags at ten fixes per second, or many tags at one fix every few seconds. TDoA scales considerably further than two-way ranging for the same airtime.
Written for engineers. Share it with one.
The hardware.
KNOWLEDGEWhich one answers your question.
SOLUTIONAcross every link layer.
SOLUTIONPlant proximity and exclusion zones.
AITiming alignment across sensors.
INDUSTRYWhere this is deployed.
Tell us the specification, the constraint and the deadline. Programmes that cross silicon, radio, embedded and AI are where Faststream is strongest.