BLE Channel Sounding vs UWB: Multipath and NLOS Accuracy Indoors
Line-of-sight demos flatter Bluetooth 6.0 channel sounding. Here is how 80 MHz versus 499.2 MHz plays out in the multipath of a real warehouse.
Hayat AminPresident of IP, Position Imaging
4 min read
With clear line of sight, Bluetooth 6.0 channel sounding ranges to within 10 to 50 cm; in multipath-heavy spaces like steel racking it slips to 1 to 3 m. UWB holds 10 to 30 cm in the same conditions because its 499.2 MHz bandwidth separates the direct path from reflections that 80 MHz of BLE spectrum blends together. If your accuracy target survives a 1 to 3 m error in your worst aisle, channel sounding's lower tag cost wins. If not, pay for UWB.
Key takeaways
- Bluetooth 6.0 channel sounding holds 10 to 50 cm in line of sight, then slips to 1 to 3 m inside steel-racked multipath.
- UWB's 499.2 MHz bandwidth resolves reflections about 60 cm apart; BLE's 80 MHz needs roughly 3.7 m of path separation.
- Match the radio to the decision it feeds: bay-level tolerance favors channel sounding cost, slot-level tolerance demands UWB.
- A UWB tag typically costs 2 to 4 times a BLE tag and burns more energy per ranging exchange.
- Solid metal blocks both radios, so anchor sightline audits matter more than the datasheet.
- Positioning patent exposure is radio-agnostic; switching from UWB to BLE does not reset your IP risk.
Why does multipath decide the channel sounding vs UWB question?
A radio's ability to tell the direct path from a reflection scales with bandwidth. A receiver can separate two arrivals only when their path lengths differ by roughly the speed of light divided by the signal bandwidth. UWB's 802.15.4z channels are 499.2 MHz wide, which resolves paths about 60 cm apart. Bluetooth 6.0 channel sounding works with about 80 MHz of 2.4 GHz spectrum, so paths need to differ by roughly 3.7 m before the math can pull them apart.
Now picture a warehouse aisle: steel racking, forklifts in motion, foil-lined packaging on every shelf. Reflections routinely arrive within 1 to 3 m of the direct path. UWB's first-path detector still locks onto the earliest pulse. Channel sounding sees a smeared phase response and reports a distance that blends the direct path with the bounce, which is why the same tag that measured 20 cm error on a conference table reads 2 m off in aisle 14. Line-of-sight demos hide this completely. Bandwidth is the whole argument.
What accuracy does Bluetooth 6.0 channel sounding actually hold indoors?
The Bluetooth 6.0 core spec, finalized in September 2024, defines two ranging methods: phase-based ranging, which steps a tone across up to 72 channels in the 2.4 GHz band, and round-trip time. Silicon shipping in 2026, including Nordic's nRF54L15 and NXP's KW47, quotes 10 to 50 cm in line of sight, and those numbers hold up on a test bench.
In reflective spaces the picture changes:
- Open ceilings with clear anchor sightlines: 0.5 to 1 m is realistic.
- Dense steel racking or ranging through inventory: expect 1 to 3 m, with occasional multi-meter outliers when a reflection dominates the phase estimate.
That still clears plenty of use cases. Bay-level pallet location, room-level equipment finding in hospitals, and secure-distance access control all live comfortably inside 1 to 2 m of error. The expensive failure mode is speccing slot-level accuracy on channel sounding, then discovering the multipath floor in month four of the pilot, after the tag order shipped. Test in your worst aisle, not the lobby.
Where does UWB keep its edge when sightlines break?
802.15.4z HRP UWB measures time of flight with pulses about 2 ns long on channels near 6.5 and 8 GHz. Because the pulse is that short, the receiver can pick the leading edge of the earliest arrival, the first path, even when stronger reflections pile in behind it. That is why Qorvo and NXP Trimension parts hold 10 to 30 cm through mild non-line-of-sight, like a tag inside a carton or behind one row of product, in conditions where channel sounding has already drifted past a meter.
Two caveats keep UWB honest. First, solid metal blocks both radios; no 8 GHz pulse ranges through a steel rack upright, so anchor placement and density still decide whether the accuracy exists at all. Second, the edge costs money. A UWB tag typically runs 2 to 4 times the bill of materials of a BLE tag, and each ranging exchange burns more energy, which shows up as shorter battery life at equal update rates. UWB pays for physics, not branding.
How do you map your facility to the right radio?
Start from the decision your coordinates feed, then walk backward to the radio.
- Bay or zone decisions, tolerance of 1 to 3 m: channel sounding wins on cost. Tags reuse BLE silicon you may already carry for telemetry, and anchors can double as gateways.
- Slot, shelf, or safety decisions, tolerance of 10 to 30 cm: UWB is the only radio choice that survives racking multipath at that tolerance.
- Mixed sites: run both. Channel sounding for the majority of assets that need bay-level answers, UWB for the forklifts and high-value tooling that need slot-level.
Then audit sightlines before you buy anything. Count how many tag positions see at least three anchors with no steel in between; if that number is low, no datasheet accuracy claim from either radio will survive contact. Ten tags of each technology in your ugliest aisle for two weeks settles the question faster than any vendor deck, and the losing radio usually loses by meters, not centimeters. Spec the radio to the decision it feeds.
Does either radio spare you positioning patent risk?
No. The ranging waveform is the standardized part; everything that turns raw distances into a product is not. Multilateration under non-line-of-sight, filtering the multipath outliers described above, fusing RF ranges with inertial or camera data, tag power scheduling, all of it sits under granted patents that predate both Bluetooth 6.0 and your product. Teams switching radios mid-project sometimes assume the exposure resets with the new chip. It does not; most positioning method claims are radio-agnostic.
This is where licensing beats rebuilding. Position Imaging licenses hundreds of granted patents in RF ranging, real-time positioning, and sensor fusion, IP that Apple, Bosch, and other major firms cite in their own filings. The portfolio includes granted RF positioning patents such as US 11,774,249 and US 12,000,947 that apply regardless of whether your tag speaks channel sounding or 802.15.4z. Pick the radio on physics and unit cost, and take the positioning methods as proven, licensed building blocks instead of a rebuild. License the ranging IP, ship the product.
Frequently asked questions
Is Bluetooth 6.0 channel sounding accurate enough for warehouse asset tracking?
For bay-level and zone-level tracking, yes. Expect 0.5 to 1 m with clear anchor sightlines and 1 to 3 m inside dense steel racking. If your workflow needs to distinguish adjacent slots or shelf positions, that multipath floor is too high and you should test UWB instead.
Why does UWB handle multipath better than BLE channel sounding?
UWB's 499.2 MHz bandwidth produces pulses around 2 ns long, so the receiver can isolate the first arriving path from reflections about 60 cm behind it. Channel sounding has roughly 80 MHz to work with, so reflections need about 3.7 m of extra path length before they can be separated from the direct signal. In a metal-heavy building, most reflections fall inside that window.
When should I choose UWB over BLE channel sounding in 2026?
Choose UWB when the decision your system feeds needs 10 to 30 cm accuracy, such as slot-level putaway, robot safety zones, or tool tracking on a line. Choose channel sounding when 1 to 3 m is acceptable and tag cost or battery life dominates the business case. Many sites justifiably run both.
What hardware supports Bluetooth channel sounding for RTLS today?
Nordic's nRF54L15 and NXP's KW47 families ship with channel sounding support, and both quote 10 to 50 cm line-of-sight ranging. On the UWB side, Qorvo and NXP Trimension parts implement 802.15.4z time-of-flight ranging. Validate any part in your own facility, since vendor accuracy figures assume line of sight.
Does switching from UWB to BLE reduce my patent exposure?
Generally no. Most granted positioning patents claim methods like multilateration, outlier filtering, and sensor fusion that apply to any ranging radio. Changing the physical layer changes your bill of materials, not your freedom to operate, which is why teams license positioning IP rather than re-engineering around it per radio.
Send us your accuracy target and your worst aisle, and we will map your tracking product to the specific granted patents that cover it.
Tell us the product. We map the exact scope, what a license covers, and how fast you can ship, all in a 20-minute call.
Book a 20-minute call