33  UWB Ranging Techniques

33.1 Start With the Story

A UWB range is the result of a timing method, not a magic number. Two-way ranging, double-sided two-way ranging, TDoA, and AoA all answer the distance question with different assumptions about clocks, anchors, tags, geometry, and traffic.

Use this chapter as a technique selection story. Start with the movement and infrastructure the system can support, then choose the ranging method whose timing evidence and failure modes can be explained.

33.2 In 60 Seconds

UWB ranging estimates distance or relative timing from very precise radio timestamps. Two-way ranging (TWR) measures a round trip between two devices and subtracts responder delay. Double-sided TWR adds a reverse exchange so clock-rate error has less influence. Time difference of arrival (TDoA) lets a tag transmit one blink that synchronized anchors timestamp, which can scale well but depends on anchor time quality. Angle of arrival (AoA) uses antenna arrays to estimate direction and is usually combined with range or TDoA evidence. In every case, the output should be treated as measurement evidence with confidence, not as an automatically exact location.

Phoebe’s Field Notes: Why Bandwidth, Not Power, Buys the Timing Resolution

Phoebe’s Why

This chapter’s TWR example subtracts timestamps as if a receiver could pin down “when the pulse arrived” to arbitrary precision, but no radio can do that – a correlator can only place an arriving pulse within a window set by how many distinct frequency components the pulse contains. A narrow channel means a smeared-out, slowly-rising pulse edge, so the best the receiver can say is “sometime in a wide window.” A wide channel means a sharp, needle-like pulse edge, so the window shrinks. That is the real reason UWB channels span hundreds of megahertz: it is not for throughput – most UWB ranging payloads are tiny – it is to buy the timing resolution this chapter’s whole technique comparison depends on. That resolution is a hard floor: no amount of averaging clock ticks or clever DS-TWR math recovers timing detail the channel bandwidth never captured in the first place.

The Derivation

A correlator resolves two timing features only if they are separated by more than about one pulse width, and a pulse confined to bandwidth \(B\) cannot be narrower than:

\[\Delta t_{min} \approx \frac{1}{B}\]

Feeding that floor through this chapter’s own one-way distance conversion:

\[d_{min} = \frac{c \times \Delta t_{min}}{2} = \frac{c}{2B}\]

Halving the achievable timing error by doubling \(B\) halves the achievable range resolution – bandwidth and ranging precision are locked together the same way frequency and wavelength were locked together for antenna sizing.

Worked Numbers: Channel 5 Versus a Narrowband Radio

• UWB channel 5 (this module’s own \(499.2\) MHz channel, from UWB Fundamentals): \(\Delta t_{min}=1/499.2\times10^{6}=2.00\) ns, so \(d_{min}=(3.00\times10^{8}\times2.00\times10^{-9})/2\) \(=0.300\) m – almost exactly the \(0.30\) m this chapter already quotes for a \(2\) ns round-trip error. That is not a coincidence picked for the example; it is the channel’s own resolution floor showing up as the chapter’s rule of thumb.
• A 20 MHz narrowband channel (catalog-typical Wi-Fi-class channel width): \(\Delta t_{min}=1/20\times10^{6}=50.0\) ns, so \(d_{min}=(3.00\times10^{8}\times50.0\times10^{-9})/2=7.50\) m – room-level at best, useless for this chapter’s \(0.25\) m robot-cell exclusion boundary.
• A 500 kHz narrowband channel (catalog-typical low-power sub-GHz channel): \(\Delta t_{min}=1/500\times10^{3}=2.00\ \mu\text{s}\), so \(d_{min}=300\) m – the timing method this chapter describes could not exist on that channel at all, regardless of clock quality or exchange count.

33.3 Learning Objectives

By the end of this chapter, you will be able to:

• Calculate a basic TWR distance from four timestamps.
• Explain why DS-TWR reduces sensitivity to clock-rate mismatch.
• Compare TWR, TDoA, and AoA by synchronization, airtime, geometry, and evidence quality.
• Identify non-line-of-sight, multipath, calibration, and clock-quality risks in UWB measurements.
• Select a ranging technique for a deployment without overclaiming accuracy, secure-ranging behavior, energy use, or scale.

33.4 Quick Check: UWB Ranging

33.5 Ranging Is Evidence

UWB does not directly output truth. It produces timestamp, channel, and geometry evidence that a positioning engine or application policy may use. A strong design records how that evidence was produced and how much confidence the system should place in it.

UWB ranging techniques compared by message pattern, clock dependency, geometry, and evidence output.

TWR A poll and response estimate round-trip time between two devices. It is simple and works without synchronized anchors, but each range consumes a bidirectional exchange.

DS-TWR A second direction or final message lets the pair reduce clock-rate error. It is useful when reply delays or clock mismatch could bias a single exchange.

TDoA A tag blink is timestamped by synchronized anchors. It shifts complexity into infrastructure and clock quality, then solves position from time differences.

AoA An antenna array estimates direction from phase or time differences across elements. It complements range or TDoA, but depends on array calibration and multipath quality.

33.6 Method Choice Is a Clock and Airtime Budget

UWB measures distance by timing radio pulses, and the named methods trade accuracy, airtime, and clock requirements differently. Single-sided two-way ranging (SS-TWR), double-sided two-way ranging (DS-TWR), and time difference of arrival (TDoA) all exploit UWB’s sub-nanosecond pulses. The design tension is that fine time resolution turns tiny clock errors into real distance errors, so each method is a different answer to “whose clock do we trust?”

Start with the conversion: light travels about 0.30 m/ns. In a two-way exchange, a timing error in the round-trip estimate is divided by two when it becomes one-way distance. A 2 ns round-trip error is therefore about 0.30 m of one-way range error: 0.30 m/ns x 2 ns / 2 = 0.30 m.

That number explains why UWB ranging is powerful but not automatic truth. If a tool cabinet only needs to decide whether a device is within about 1.5 m, a well-calibrated DS-TWR exchange with confidence fields may be enough. If a robot cell needs a 0.25 m exclusion boundary, the same raw range is not enough unless the system also proves antenna-delay calibration, first-path quality, line-of-sight evidence, freshness, and geometry. The method name says how evidence was collected; it does not say the evidence is safe to act on.

A useful range record carries more than distance. It should include method, anchors or peer, raw timing age, calibration version, channel quality, confidence class, and the policy action that consumed it. A log entry such as “DS-TWR, 3.2 m, high confidence, 45 ms old, calibration v4, accepted” is operational evidence. A bare “3.2 m” is a number with hidden assumptions.

33.7 Two-Way Ranging

Single-sided TWR uses four timestamps:

1. The initiator transmits a poll at (T_1).
2. The responder receives it at (T_2).
3. The responder transmits a response at (T_3).
4. The initiator receives the response at (T_4).

The responder delay is (T_3 - T_2). The initiator’s elapsed time is (T_4 - T_1). The round-trip flight time estimate is:

\[ RTT = (T_4 - T_1) - (T_3 - T_2) \]

Distance is:

\[ d = \frac{c \times RTT}{2} \]

where (c) is the speed of light in air for the required engineering precision.

Two-way ranging times a poll-and-response exchange (T1 to T4) to compute time-of-flight and convert it to a centimeter-level distance.

33.7.1 Worked TWR Example

Suppose a tag sends a poll at (T_1 = 0) ns and receives the response at (T_4 = 220) ns. The responder records (T_2 = 10) ns and (T_3 = 210) ns.

• Initiator elapsed time: (T_4 - T_1 = 220) ns.
• Responder delay: (T_3 - T_2 = 200) ns.
• Round-trip flight time: (220 - 200 = 20) ns.
• One-way flight time: (20 / 2 = 10) ns.
• Distance using (c ) m/ns: (10 = 3.0) m.

This calculation is useful only if timestamp calibration, antenna delay, clock behavior, and channel quality are good enough for the decision that will use the result.

33.8 Try Two-Way Ranging in the Workbench

33.9 Double-Sided TWR

Single-sided TWR assumes the responder delay measured on one device’s clock is directly comparable with the elapsed time measured on the other device’s clock. Small clock-rate differences can matter when reply delays are long or timing quality is weak. DS-TWR adds a reverse exchange or a final message so the ranging engine has timing measurements from both sides.

DS-TWR does not make measurements perfect. It reduces one error source: clock-rate mismatch between the devices. The result still needs antenna-delay calibration, first-path review, and quality checks.

Use DS-TWR when:

• The exchange is peer-to-peer or small-scale.
• A single-sided exchange may be biased by clock-rate mismatch.
• A bidirectional exchange is acceptable for latency and airtime.
• The application needs range confidence, not just a nearby/not-nearby signal.

33.10 Time Difference of Arrival

TDoA changes the architecture. The tag transmits a blink, and multiple anchors timestamp the arrival. The positioning engine compares arrival-time differences, so it can estimate a location from hyperbolic constraints. The tag does not need to range separately with every anchor, but the anchor network must maintain time alignment and good geometry.

For two anchors, the measured time difference gives a distance-difference constraint:

\[ d_2 - d_1 = c \times (t_2 - t_1) \]

The positioning engine combines several such constraints with anchor coordinates, clock-quality estimates, and filtering. Poor synchronization directly becomes bad range-difference evidence. Poor geometry can amplify otherwise small timing errors.

UWB positioning architectures: two-way ranging versus TDoA

Use TDoA when:

• Many tags need periodic position evidence.
• The deployment can support synchronized anchor infrastructure.
• The system can monitor clock quality, anchor health, and geometry.
• A central engine can reject, down-weight, or flag weak measurements.

Avoid treating TDoA as simply “lower power” or “larger scale.” Those outcomes depend on update rates, channel planning, tag behavior, anchor density, synchronization quality, and application tolerance for stale or uncertain evidence.

33.11 Angle of Arrival

AoA estimates direction rather than distance. A UWB receiver with an antenna array compares phase or timing evidence across known antenna positions. Direction evidence can reduce ambiguity or improve geometry, especially when combined with TWR or TDoA. It also creates new failure modes: array calibration, device orientation, polarization, body blockage, and reflections can bias the angle estimate.

Use AoA as a complement when:

• A direction estimate improves the application decision.
• The anchor can support calibrated antenna arrays.
• Multipath and mounting orientation are included in the test plan.
• The positioning engine can attach confidence to angle evidence.

33.12 Measurement Quality

Every ranging technique needs a quality gate before an application acts. Good systems do not only store a distance or coordinate; they store enough evidence to explain why the measurement was accepted, down-weighted, retried, or rejected.

UWB measurement quality path from raw timestamps and channel impulse response through calibration, geometry, confidence, and application action.

Review these quality factors:

• Clock quality: TDoA depends on anchor synchronization; TWR depends on stable local timestamping and bounded reply behavior.
• Calibration: Antenna delay and device-specific offsets must be measured or controlled.
• First-path evidence: Channel impulse response helps separate the earliest useful path from stronger reflections.
• Geometry: Anchor layout can amplify timing errors into larger position errors.
• Motion and latency: A good range can still be stale when the tag or controlled object is moving.
• Policy fit: A measurement that is good enough for room-level context may not be good enough for access or safety behavior.

33.13 Technique Selection

Choose the method by the evidence requirement, not by a headline accuracy number.

Peer-to-peer or small set Start with TWR or DS-TWR. Confirm latency, airtime, reply timing, and confidence under real mounting conditions.

Many managed tags Consider TDoA if synchronized anchors, monitoring, calibration, and backend processing are part of the deployment plan.

Direction matters Add AoA when the application benefits from bearing evidence, but test array calibration and reflection sensitivity.

High-consequence decisions Require confidence, freshness, fallback behavior, logs, and independent controls where a wrong position could cause harm.

33.13.1 Example Selection Reasoning

A tool cabinet only needs to know whether one authorized device is near the correct side of the cabinet at checkout time. DS-TWR can fit because the interaction is local, bidirectional, and low population. A factory-wide asset map with many tags may fit TDoA if the site can maintain synchronized anchors and confidence monitoring. A robot docking system might combine range with AoA if direction helps resolve approach angle. None of these choices is complete until the installed environment proves the evidence quality.

33.14 Capacity and Failure-Mode Review

The practical split is two-way versus one-way timing. TWR needs a back-and-forth with each anchor, which costs airtime per tag but needs no synchronized infrastructure. TDoA reverses that model: the tag blinks once and the fixed anchors, sharing a common clock, compute position from arrival-time differences.

Method	Exchange	Clock requirement	Trade
SS-TWR	Poll -> response; one side times the round trip	No shared clock, but sensitive to inter-device clock offset	Simple and fewer messages; usually least accurate
DS-TWR	Poll -> response -> final; both sides time a round trip	No shared clock; the math cancels much first-order clock offset	More accurate; one extra message and more energy
TDoA	Tag sends one blink; many anchors timestamp arrival	Anchors must be tightly synchronized	Tag transmits once, so it scales to many tags when infrastructure is healthy

Do the capacity arithmetic before choosing. A small tool-checkout system with 12 active tags, 4 anchors per decision, and 2 Hz updates asks for about 12 x 4 x 2 = 96 tag-anchor ranging decisions per second. If DS-TWR uses three messages for each decision, the message pressure is roughly 96 x 3 = 288 ranging messages per second before retries and management traffic. The same population using TDoA at 2 Hz asks the tags for about 12 x 2 = 24 blinks per second, but the anchors must maintain synchronization and the backend must reject weak geometry.

Acceptance testing should mirror the decision. For TWR, test known distances with the installed antenna orientation and record responder delay, channel impulse response first-path quality, and rejected measurements. For TDoA, test known points inside and near the edge of the anchor hull, then record anchor time status and residual error. For AoA, include a bearing sanity check: a 10 degree angle error at 5 m is about 5 x tan(10 degrees) = 0.88 m sideways uncertainty. That is fine for “which aisle” and weak for a tight safety zone.

The practitioner rule is simple: choose the method whose failure mode you can monitor. TWR failure often appears as airtime pressure, reply timing, or non-line-of-sight bias. TDoA failure often appears as clock drift, bad geometry, or stale anchor coordinates. AoA failure often appears as array calibration, mounting, or reflections. A design review that names those checks is stronger than one that only cites an accuracy headline.

33.15 Clock Drift Error Budget

In SS-TWR, device A sends a poll, device B replies after a known processing delay, and A computes distance from the round-trip time minus B’s reply delay. The catch is that the reply delay is measured on B’s clock while the round trip is measured on A’s clock. Those crystals never run at exactly the same frequency. That frequency offset multiplies the reply delay into a ranging error, and reply delays are long enough that small parts-per-million offsets become centimeters.

DS-TWR reduces this by adding a third message so each device measures a round trip. Combining the two round-trip measurements cancels much of the first-order clock-offset error. The cost is one more message and more energy per range. The choice is concrete: SS-TWR when the fewest messages matter and more error is acceptable; DS-TWR when accuracy must hold despite clock drift; TDoA when tag scale and battery life matter more than avoiding anchor synchronization.

Put numbers on the drift. Suppose the responder waits 100 microseconds before sending its reply and its clock differs by 20 ppm from the initiator. The apparent reply-delay error is about 100 microseconds x 20 / 1,000,000 = 2 ns. In a single-sided range, that becomes about 0.30 m/ns x 2 ns / 2 = 0.30 m of distance bias. If the reply delay stretches to 1 ms, the same 20 ppm offset becomes 20 ns, or roughly 3 m of distance bias before multipath, antenna-delay, or timestamp-noise effects.

DS-TWR still leaves the rest of the error budget. Antenna delay can shift timestamps, non-line-of-sight paths can arrive late, and filtering can publish a fresh-looking range from stale evidence. TDoA has a different budget: if anchor synchronization is off by 0.5 ns, the distance-difference evidence is biased by about 0.15 m. With a geometry factor of 2, that can become about 0.30 m of position uncertainty before environment bias.

Engineering review should attach method, timing age, calibration, channel evidence, geometry, and confidence to every accepted range or coordinate. The distance is only trustworthy when the evidence trail matches the application consequence.

33.16 Common Pitfalls

• Claiming automatic centimeter accuracy. UWB can support fine timing evidence, but installed accuracy depends on line of sight, calibration, geometry, clock quality, antenna placement, and filtering.
• Using TDoA without clock evidence. Time-difference systems are only as good as anchor synchronization, timestamping, and drift monitoring.
• Ignoring non-line-of-sight bias. A reflected path can arrive later than the direct path and make the range look too long. Use channel evidence and consistency checks.
• Optimizing update rate before airtime. More ranges per second can increase collisions, stale processing, and energy use. Check packet timing and channel load.
• Turning range into authorization. Secure ranging can strengthen proximity evidence, but authorization, freshness, confidence, boundary policy, and fallback still decide the action.

33.17 Knowledge Check

33.17.1 Quick Check: TWR Timing

33.17.2 Matching Quiz: Ranging Evidence

33.17.3 Ordering Quiz: TDoA Measurement Review

33.17.4 Quick Check: Clock Offset

33.18 Summary

• TWR estimates distance from a bidirectional exchange and does not require synchronized anchors.
• DS-TWR reduces sensitivity to clock-rate mismatch but still needs channel, calibration, and confidence review.
• TDoA can support managed multi-tag systems when synchronized anchors and backend quality monitoring are available.
• AoA adds direction evidence, usually as a complement to range or time-difference measurements.
• UWB ranging results should be handled as evidence with confidence, freshness, geometry, and fallback rules.

33.19 Key Takeaway

UWB ranging quality depends on timing method, anchor geometry, line of sight, multipath, calibration, and validation against the required accuracy.

33.20 Concept Relationships

• Wide bandwidth supports precise timestamp evidence.
• TWR and DS-TWR turn round-trip timing into range evidence.
• TDoA turns synchronized arrival-time differences into position constraints.
• AoA adds direction evidence through antenna geometry.
• CIR review helps distinguish first-path evidence from reflections.
• Application policy decides whether evidence is strong enough to accept, retry, down-weight, or fall back.

33.21 References

• IEEE 802.15.4z-2020: enhanced UWB ranging amendment.
• IEEE 802.15.4: low-rate wireless networks and UWB PHY history.
• FiRa Consortium technical resources: UWB ranging, MAC, and interoperability guidance.
• Decawave application notes: UWB two-way ranging, channel impulse response, and antenna delay calibration.
• ETSI EN 302 065 and FCC UWB operation rules: regional spectrum constraints for UWB devices.

33.22 What’s Next

UWB Positioning Systems Turn range, time-difference, and angle evidence into deployed location systems.

UWB Applications and Security Apply range confidence to access, asset tracking, safety, privacy, and operations.

UWB Fundamentals Review wide bandwidth, timing evidence, channel impulse response, and spectrum constraints.

Ultra-Wideband Place ranging techniques in the broader UWB technology landscape.