Bluetooth Channel Sounding vs UWB: Why 80 MHz Caps Your Accuracy

What accuracy does Bluetooth 6.0 Channel Sounding hardware deliver in 2026?

Bluetooth 6.0, published in September 2024, added Channel Sounding: the radio hops across up to 72 channels in the 2.4 GHz band, measures phase at each one, and cross-checks the result with round-trip timing. Two years later the silicon is real. Nordic's nRF54L15 and NXP's KW47 ship Channel Sounding in production, and both vendors demonstrate 30 to 50 cm ranging in line of sight.

Those demos are honest, and they are also the best case. Put the same radios in a warehouse aisle with steel racking, shrink-wrapped pallets, and forklifts crossing the link, and repeatable Channel Sounding accuracy sits between 1 and 3 m. UWB hardware measured in the same conditions, Qorvo DW3000-class chips or NXP Trimension modules running IEEE 802.15.4z two-way ranging, holds 10 to 30 cm.

That is the 2026 scoreboard. Channel Sounding is a real ranging technology, not an RSSI guess, but it degrades roughly 3x to 6x when the environment turns hostile, while UWB barely moves. Datasheets record best days, not worst aisles.

Why does 80 MHz of spectrum cap what Channel Sounding can resolve?

Ranging accuracy in multipath comes down to bandwidth. A radio can separate two signal paths only when their arrival times differ by more than roughly one over the bandwidth, which works out to a path-resolution distance of about c divided by twice the bandwidth. Bluetooth has about 80 MHz to work with, 2400 to 2483.5 MHz, so reflections arriving within roughly 1.9 m of the direct path blur into a single measurement. UWB's 499.2 MHz channels at 6.5 and 8 GHz resolve paths about 30 cm apart, so a bounce off a rack shows up as a separate, rejectable pulse instead of a corrupted estimate.

Phase-based ranging is the clever part of Channel Sounding. By fitting the phase slope across all the sounded tones, it beats the raw 1.9 m figure in clean air, which is exactly how vendors reach 30 to 50 cm in line of sight. But when a strong reflection lands inside the resolution window, the phases of both paths add before the fit runs, and the estimate settles somewhere between them. No firmware update changes that arithmetic. Bandwidth is physics, and physics does not negotiate.

Where does Channel Sounding beat UWB anyway?

Three places, and they matter.

• Tag cost. Channel Sounding runs on the same SoC that already runs the tag's Bluetooth stack. A UWB tag needs a second radio and a second antenna, and finished UWB asset tags typically price at two to four times a comparable BLE tag. Across 20,000 pallets, that gap is the budget.
• Power. A Channel Sounding exchange costs more energy than a plain advertising beacon, but a CR2032 coin cell still supports years of service at a few rangings per minute. UWB's higher peak transmit current pushes designs toward larger cells or lower duty cycles.
• Upgrade path. Sites already running BLE gateways for beaconing can move to Channel Sounding by swapping endpoints, not rewiring the ceiling.

The advantage Channel Sounding does not have yet is phones. UWB has shipped in Apple flagships since the iPhone 11 in 2019 and in Samsung Galaxy flagships since 2020, while handsets with Channel Sounding silicon are only starting to appear. For 2026 deployments, plan on dedicated readers either way. Good enough at half the price usually wins.

How should you choose for asset tracking in 2026?

Start from the decision the position feeds, not the radio.

• The system answers "which zone, which aisle": Channel Sounding. Real-world accuracy of 1 to 3 m resolves aisles spaced 3 m apart, and tag cost lets you cover every pallet instead of a sample.
• The system answers "which shelf slot, which side of the forklift": UWB. Sub-30 cm near metal is not a Channel Sounding outcome in 2026, and a safety function cannot run on a radio that blurs by meters when a truck drives through the link.
• Secure ranging gates a door or a vehicle: UWB again. The 802.15.4z scrambled timestamp sequence was designed against relay attacks and backs the Car Connectivity Consortium's Digital Key 3.0.
• Mixed fleet: put Channel Sounding on the bulk of assets and UWB on the fraction that needs precision, solved in one position engine.

Before committing a BOM, run both radios for two weeks in your worst aisle, not your lab.

Match the radio to the decision it feeds.

What does the choice mean for patent exposure?

Both stacks sit on two decades of granted RF ranging patents: phase-based distance estimation, two-way timing exchanges, multi-anchor position solving, and the multipath filtering that separates a demo from a product. The radio you pick changes your BOM more than it changes that exposure, because the value, and the claims, concentrate in the system layer above the chip.

That layer is buildable, but it is a multi-year effort. Position Imaging licenses hundreds of granted patents in radio-frequency ranging, real-time positioning, and computer vision, IP cited by Apple, Bosch, and other major firms, including US 11,774,249, US 12,079,006, US 12,066,561, and US 12,000,947. Teams license the portfolio to ship tracking products in months with freedom to operate, instead of re-deriving multipath handling from scratch and hoping the claim charts come back clean. License the ranging problem, spend your engineers elsewhere.